Back to the Finder

gh-aw

Open-source desktop app for Mac, Windows and Linux. Agents run on your computer, on your own server or in the vendor's cloud.

Product websitegithub.com Compare with others Missing or wrong data?

gh-aw product interface, supplied by its publisher
Desktop app
Mac, Windows and Linux
Browser
No
Phone
No
Command line
Yes
Agents
Claude Code, Codex, Cursor Agent, Gemini CLI, OpenCode, Pi, Aider and Antigravity
Your subscription
Works with your own plan
Agents run
On your computer, on your own server and in the vendor's cloud
Laptop closed
Keeps going
Isolation
Container and VM
Parallel agents
Yes
Scheduled runs
Yes
Coordinator agent
Yes
Built-in issues
Yes
Pull requests
Yes
Review screen
No
Team
No
Self-hosting
Not listed
Licence
MIT
Price from
Free
How it charges
No paid plan and bring your own key
Paid plans from
No paid plan

Receipts

Checked
· oldest answer 2026-11-08
Findings
34 published · 22 cross-checked by a second agent · read them all
Sources
26 public pages on 1 site: github.com
Screenshot
github.com, retrieved · first party repository · shows appearance, not capability
Method
Same rules for all 103 researched ADEs. How we check · Report wrong data

What it does

Where you use it

gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

Source and scope for Where you use it

Agents it runs

Built-in engines and custom imported definitions are separate. Public experimental samples launch Cursor Agent, OpenCode and Aider without compatibility guarantees; these are runnable integrations, not supported built-in engines.

Source and scope for Agents it runs

Accounts and subscriptions

A personal Copilot subscription can use COPILOT_GITHUB_TOKEN. Claude subscription OAuth and ChatGPT subscription login are unsupported. Claude and Codex support API credentials and provider-specific inference routes.

Source and scope for Accounts and subscriptions

Where agents run

Compiled workflows execute on GitHub Actions runners, including user-controlled Linux runners. A runner on the user computer executes there; other self-hosted runners and GitHub-hosted runners remain separate options.

Source and scope for Where agents run

How work is kept apart

AWF Docker isolation is the default; opt-in Cloud Hypervisor/NVX preview runtimes use KVM microVMs on eligible runners. The compiler does not provision Git worktrees.

Source and scope for How work is kept apart

Parallel work

Workflows on distinct issues/PRs/branches may run concurrently; default per-engine groups serialize each engine.

Source and scope for Parallel work

Scheduled runs

Schedule frontmatter can run a recurring workflow such as a daily report. GitHub events can start automation without a human mentioning the agent on every run. Schedules execute through GitHub Actions rather than a local browser or desktop scheduler.

Source and scope for Scheduled runs

Git and pull requests

create-pull-request safe outputs propose code in a separate permission-scoped job. Workflow examples include CI-failure investigation; automatically triggering CI from generated PRs may require an additional token.

Source and scope for Git and pull requests

Licence and source

The gh-aw compiler and runtime repository use MIT.

Source and scope for Licence and source

Price

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

Source and scope for Price

JSON for agentsMarkdown for agents

Report wrong dataSuggest a missing ADE

Typed filters

Every value links to its receipt below. Read plan and surface restrictions before using a filter to decide.

Typed filter values
FilterValue
Desktop OS: macOSYes
Desktop OS: WindowsYes
Desktop OS: LinuxYes
Web clientNo
Phone access: Native iOSNo
Phone access: Native AndroidNo
Phone access: Phone browserNo
CLIYes cross-checked
Coding agents: Claude CodeYes
Coding agents: CodexYes
Coding agents: Cursor AgentYes
Coding agents: Gemini CLIYes
Coding agents: OpenCodeYes
Coding agents: PiYes
Coding agents: GrokUnknown
Coding agents: AiderYes
Coding agents: AntigravityYes
Use my subscriptionYes cross-checked
Where agents run: On your computerYes
Where agents run: On your own serverYes
Where agents run: In the vendor's cloudYes
Workspace isolation: Git worktreeNo
Workspace isolation: ContainerYes
Workspace isolation: VM or microVMYes
Parallel runsYes cross-checked
CoordinatorYes
Native issue trackingYes cross-checked
Unattended triggersYes cross-checked
Shared promptingNo
Team rolesNo
SSONo
Pull request workflowYes cross-checked
Review interfaceNo
Product license: Apache-2.0No. MIT cross-checked
Product license: MITYes. MIT cross-checked
Product license: GPL-3.0No. MIT cross-checked
Product license: GPL-3.0-or-laterNo. MIT cross-checked
Product license: AGPL-3.0No. MIT cross-checked
Product license: AGPL-3.0-or-laterNo. MIT cross-checked
Product license: Elastic License 2.0No. MIT cross-checked
Product license: Custom termsNo. MIT cross-checked
Product license: ProprietaryNo. MIT cross-checked
Self-hostingUnknown
Telemetry opt-outYes
Free accessYes cross-checked
Starting price: FreeYes. Free cross-checked
Starting price: Up to $20/monthNo. Free cross-checked
Starting price: $21–40/monthNo. Free cross-checked
Starting price: $41–100/monthNo. Free cross-checked
Starting price: Over $100/monthNo. Free cross-checked
Starting price: Contact for pricingNo. Free cross-checked
Pricing model: Free, no paid planYes
Pricing model: Free planNo
Pricing model: SubscriptionNo
Pricing model: Per seatNo
Pricing model: Usage-basedNo
Pricing model: Per machine or computeNo
Pricing model: Bring your own keyYes cross-checked
Pricing model: One-time purchaseNo
Paid plans from: No paid planYes. No paid plan
Paid plans from: $10 or less/monthNo. No paid plan
Paid plans from: $11–20/monthNo. No paid plan
Paid plans from: $21–40/monthNo. No paid plan
Paid plans from: $41–100/monthNo. No paid plan
Paid plans from: Over $100/monthNo. No paid plan
Paid plans from: Contact salesNo. No paid plan
Lifecycle status: ActiveUnknown
Lifecycle status: SunsettingUnknown
Keeps going with laptop closedYes
Open-source licenseYes. MIT cross-checked
Any desktop appYes
Paid plan up to $20/monthNo. No paid plan
Paid plan up to $40/monthNo. No paid plan
Any phone clientNo

Filter receipts

Desktop OS: macOS: Yes

Desktop OS: macOS

Typed value: Yes.

gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

  • Desktop os: Native release extension binaries cover macOS, Windows and Linux; Windows installer uses Git Bash/MSYS/Cygwin.
  • Mobile apps: No native iOS/Android application; Android Termux CLI installation is not a native phone app.
  • Browser: GitHub workflow pages are the external platform UI; the extension does not supply an ADE browser client.
  • Mobile browser: No product-owned phone-browser ADE interface.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Desktop OS: Windows: Yes

Desktop OS: Windows

Typed value: Yes.

gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

  • Desktop os: Native release extension binaries cover macOS, Windows and Linux; Windows installer uses Git Bash/MSYS/Cygwin.
  • Mobile apps: No native iOS/Android application; Android Termux CLI installation is not a native phone app.
  • Browser: GitHub workflow pages are the external platform UI; the extension does not supply an ADE browser client.
  • Mobile browser: No product-owned phone-browser ADE interface.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Desktop OS: Linux: Yes

Desktop OS: Linux

Typed value: Yes.

gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

  • Desktop os: Native release extension binaries cover macOS, Windows and Linux; Windows installer uses Git Bash/MSYS/Cygwin.
  • Mobile apps: No native iOS/Android application; Android Termux CLI installation is not a native phone app.
  • Browser: GitHub workflow pages are the external platform UI; the extension does not supply an ADE browser client.
  • Mobile browser: No product-owned phone-browser ADE interface.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Web client: No

Web client

Typed value: No.

gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

  • Desktop os: Native release extension binaries cover macOS, Windows and Linux; Windows installer uses Git Bash/MSYS/Cygwin.
  • Mobile apps: No native iOS/Android application; Android Termux CLI installation is not a native phone app.
  • Browser: GitHub workflow pages are the external platform UI; the extension does not supply an ADE browser client.
  • Mobile browser: No product-owned phone-browser ADE interface.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Phone access: Native iOS: No

Phone access: Native iOS

Typed value: No.

gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

  • Desktop os: Native release extension binaries cover macOS, Windows and Linux; Windows installer uses Git Bash/MSYS/Cygwin.
  • Mobile apps: No native iOS/Android application; Android Termux CLI installation is not a native phone app.
  • Browser: GitHub workflow pages are the external platform UI; the extension does not supply an ADE browser client.
  • Mobile browser: No product-owned phone-browser ADE interface.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Phone access: Native Android: No

Phone access: Native Android

Typed value: No.

gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

  • Desktop os: Native release extension binaries cover macOS, Windows and Linux; Windows installer uses Git Bash/MSYS/Cygwin.
  • Mobile apps: No native iOS/Android application; Android Termux CLI installation is not a native phone app.
  • Browser: GitHub workflow pages are the external platform UI; the extension does not supply an ADE browser client.
  • Mobile browser: No product-owned phone-browser ADE interface.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Phone access: Phone browser: No

Phone access: Phone browser

Typed value: No.

gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

  • Desktop os: Native release extension binaries cover macOS, Windows and Linux; Windows installer uses Git Bash/MSYS/Cygwin.
  • Mobile apps: No native iOS/Android application; Android Termux CLI installation is not a native phone app.
  • Browser: GitHub workflow pages are the external platform UI; the extension does not supply an ADE browser client.
  • Mobile browser: No product-owned phone-browser ADE interface.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

CLI: Yes cross-checked

CLI

Typed value: Yes.

gh aw compile, logs and audit provide workflow authoring and inspection operations. gh aw mcp-server exposes workflow operations, read-only unless write tools are enabled.

  • Cli: gh aw compile, logs and audit provide workflow authoring and inspection operations.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows README.md

    Finding: The gh aw extension compiles Markdown/YAML repository automation into GitHub Actions workflows with configurable agents, read-only sandbox defaults and separate safe-output jobs.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

Coding agents: Claude Code: Yes

Coding agents: Claude Code

Typed value: Yes.

Built-in engines and custom imported definitions are separate. Public experimental samples launch Cursor Agent, OpenCode and Aider without compatibility guarantees; these are runnable integrations, not supported built-in engines.

  • Session integration: Compiled workflows launch the configured native engine with declared tools and sandbox policy.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  • The released Agy engine is experimental, Linux x64 only, and supports Gemini API keys rather than Google WIF, ADC or Vertex AI. It cannot disable native shell, editing or web tools.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/engines.md

    Finding: Built-in engines include Copilot, Claude Code, Codex, deprecated Gemini CLI and Pi. Experimental durable mode checkpoints native sessions in artifacts; max-turns and AI-credit limits vary by engine.

Coding agents: Codex: Yes

Coding agents: Codex

Typed value: Yes.

Built-in engines and custom imported definitions are separate. Public experimental samples launch Cursor Agent, OpenCode and Aider without compatibility guarantees; these are runnable integrations, not supported built-in engines.

  • Session integration: Compiled workflows launch the configured native engine with declared tools and sandbox policy.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  • The released Agy engine is experimental, Linux x64 only, and supports Gemini API keys rather than Google WIF, ADC or Vertex AI. It cannot disable native shell, editing or web tools.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/engines.md

    Finding: Built-in engines include Copilot, Claude Code, Codex, deprecated Gemini CLI and Pi. Experimental durable mode checkpoints native sessions in artifacts; max-turns and AI-credit limits vary by engine.

Coding agents: Cursor Agent: Yes

Coding agents: Cursor Agent

Typed value: Yes.

Built-in engines and custom imported definitions are separate. Public experimental samples launch Cursor Agent, OpenCode and Aider without compatibility guarantees; these are runnable integrations, not supported built-in engines.

  • Harnesses: Published experimental custom-engine samples launch Cursor Agent, OpenCode 2 and Aider; existing built-in engines retain their own scope.
  • Custom adapters: Compiler accepts imported custom engine definitions, including the shipped samples.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  • The released Agy engine is experimental, Linux x64 only, and supports Gemini API keys rather than Google WIF, ADC or Vertex AI. It cannot disable native shell, editing or web tools.
  1. gh-aw: .github/workflows/shared/cursor.md

    Finding: The repository ships publicly importable experimental engine definitions that install/launch Cursor Agent, OpenCode 2 and Aider. They are explicitly unsupported examples for integration authors without a compatibility or maintenance commitment, rather than built-in supported engines. Custom engine definitions are accepted by the compiler.

Coding agents: Gemini CLI: Yes

Coding agents: Gemini CLI

Typed value: Yes.

Built-in engines and custom imported definitions are separate. Public experimental samples launch Cursor Agent, OpenCode and Aider without compatibility guarantees; these are runnable integrations, not supported built-in engines.

  • Session integration: Compiled workflows launch the configured native engine with declared tools and sandbox policy.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  • The released Agy engine is experimental, Linux x64 only, and supports Gemini API keys rather than Google WIF, ADC or Vertex AI. It cannot disable native shell, editing or web tools.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/engines.md

    Finding: Built-in engines include Copilot, Claude Code, Codex, deprecated Gemini CLI and Pi. Experimental durable mode checkpoints native sessions in artifacts; max-turns and AI-credit limits vary by engine.

Coding agents: OpenCode: Yes

Coding agents: OpenCode

Typed value: Yes.

Built-in engines and custom imported definitions are separate. Public experimental samples launch Cursor Agent, OpenCode and Aider without compatibility guarantees; these are runnable integrations, not supported built-in engines.

  • Harnesses: Published experimental custom-engine samples launch Cursor Agent, OpenCode 2 and Aider; existing built-in engines retain their own scope.
  • Custom adapters: Compiler accepts imported custom engine definitions, including the shipped samples.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  • The released Agy engine is experimental, Linux x64 only, and supports Gemini API keys rather than Google WIF, ADC or Vertex AI. It cannot disable native shell, editing or web tools.
  1. gh-aw: .github/workflows/shared/cursor.md

    Finding: The repository ships publicly importable experimental engine definitions that install/launch Cursor Agent, OpenCode 2 and Aider. They are explicitly unsupported examples for integration authors without a compatibility or maintenance commitment, rather than built-in supported engines. Custom engine definitions are accepted by the compiler.

Coding agents: Pi: Yes

Coding agents: Pi

Typed value: Yes.

Built-in engines and custom imported definitions are separate. Public experimental samples launch Cursor Agent, OpenCode and Aider without compatibility guarantees; these are runnable integrations, not supported built-in engines.

  • Session integration: Compiled workflows launch the configured native engine with declared tools and sandbox policy.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  • The released Agy engine is experimental, Linux x64 only, and supports Gemini API keys rather than Google WIF, ADC or Vertex AI. It cannot disable native shell, editing or web tools.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/engines.md

    Finding: Built-in engines include Copilot, Claude Code, Codex, deprecated Gemini CLI and Pi. Experimental durable mode checkpoints native sessions in artifacts; max-turns and AI-credit limits vary by engine.

Coding agents: Aider: Yes

Coding agents: Aider

Typed value: Yes.

Built-in engines and custom imported definitions are separate. Public experimental samples launch Cursor Agent, OpenCode and Aider without compatibility guarantees; these are runnable integrations, not supported built-in engines.

  • Harnesses: Published experimental custom-engine samples launch Cursor Agent, OpenCode 2 and Aider; existing built-in engines retain their own scope.
  • Custom adapters: Compiler accepts imported custom engine definitions, including the shipped samples.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  • The released Agy engine is experimental, Linux x64 only, and supports Gemini API keys rather than Google WIF, ADC or Vertex AI. It cannot disable native shell, editing or web tools.
  1. gh-aw: .github/workflows/shared/cursor.md

    Finding: The repository ships publicly importable experimental engine definitions that install/launch Cursor Agent, OpenCode 2 and Aider. They are explicitly unsupported examples for integration authors without a compatibility or maintenance commitment, rather than built-in supported engines. Custom engine definitions are accepted by the compiler.

Coding agents: Antigravity: Yes

Coding agents: Antigravity

Typed value: Yes.

Built-in engines and custom imported definitions are separate. Public experimental samples launch Cursor Agent, OpenCode and Aider without compatibility guarantees; these are runnable integrations, not supported built-in engines.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  • The released Agy engine is experimental, Linux x64 only, and supports Gemini API keys rather than Google WIF, ADC or Vertex AI. It cannot disable native shell, editing or web tools.
  1. gh-aw Antigravity support at v1.2.2

    Finding: gh-aw v1.2.2 ships the native experimental Agy engine and documents engine: agy running Antigravity CLI in GitHub Actions.

  2. gh-aw Antigravity support at v1.2.2

    Finding: gh-aw v1.2.2 ships the native experimental Agy engine and documents engine: agy running Antigravity CLI in GitHub Actions.

Use my subscription: Yes cross-checked

Use my subscription

Typed value: Yes.

A personal Copilot subscription can use COPILOT_GITHUB_TOKEN. Claude subscription OAuth and ChatGPT subscription login are unsupported. Claude and Codex support API credentials and provider-specific inference routes.

  • Agent subscriptions: A personal Copilot subscription can use COPILOT_GITHUB_TOKEN. Claude subscription OAuth and ChatGPT subscription login are unsupported.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/engines/copilot.md

    Finding: Personal Copilot subscriptions can be used with COPILOT_GITHUB_TOKEN, separate from organization-billed inference.

Where agents run: On your computer: Yes

Where agents run: On your computer

Typed value: Yes.

Compiled workflows execute on GitHub Actions runners, including user-controlled Linux runners. A runner on the user computer executes there; other self-hosted runners and GitHub-hosted runners remain separate options.

  • Local: A user-managed Linux runner on the user computer executes the compiled agent job there; compilation by the CLI alone is not agent execution.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Where agents run: On your own server: Yes

Where agents run: On your own server

Typed value: Yes.

Compiled workflows execute on GitHub Actions runners, including user-controlled Linux runners. A runner on the user computer executes there; other self-hosted runners and GitHub-hosted runners remain separate options.

  • Byo remote: Self-hosted Linux container-capable runners can execute the compiled workflow.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Where agents run: In the vendor's cloud: Yes

Where agents run: In the vendor's cloud

Typed value: Yes.

Compiled workflows execute on GitHub Actions runners, including user-controlled Linux runners. A runner on the user computer executes there; other self-hosted runners and GitHub-hosted runners remain separate options.

  • Managed cloud: On GitHub-hosted Actions runners the agent runs in an AWF Docker container by default.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/sandbox.md

    Finding: The default docker runtime runs the agent in a rootless AWF container with network isolation; preview cloud-hypervisor and nvx runtimes use KVM microVMs. Filesystem mounts and outbound network allowlists are configurable. The page does not itself describe the GitHub runner as a VM.

  2. gh-aw: Security architecture

    Finding: The default agent runtime uses Docker container isolation on GitHub-hosted or Linux self-hosted runners, with the Squid proxy as the only egress path. Eligible GitHub-hosted Ubuntu x86_64 KVM runners can instead place the agent in a Cloud Hypervisor microVM (preview).

Workspace isolation: Git worktree: No

Workspace isolation: Git worktree

Typed value: No.

AWF Docker isolation is the default; opt-in Cloud Hypervisor/NVX preview runtimes use KVM microVMs on eligible runners. The compiler does not provision Git worktrees.

  • Worktree: Checkout/branch steps use the workflow workspace, not a product worktree isolation option.
  • MicroVM runtimes are previews limited to eligible KVM runners; the default is a Docker container.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Workspace isolation: Container: Yes

Workspace isolation: Container

Typed value: Yes.

AWF Docker isolation is the default; opt-in Cloud Hypervisor/NVX preview runtimes use KVM microVMs on eligible runners. The compiler does not provision Git worktrees.

  • Container: The default docker runtime runs the agent in a rootless AWF container.
  • Vm: Preview cloud-hypervisor (GitHub-hosted Ubuntu x86_64 KVM runners) and nvx runtimes place the agent in a KVM microVM.
  • Network: AWF network isolation sends egress only through the proxy to configured domains.
  • MicroVM runtimes are previews limited to eligible KVM runners; the default is a Docker container.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/sandbox.md

    Finding: The default docker runtime runs the agent in a rootless AWF container with network isolation; preview cloud-hypervisor and nvx runtimes use KVM microVMs. Filesystem mounts and outbound network allowlists are configurable. The page does not itself describe the GitHub runner as a VM.

  2. gh-aw: Security architecture

    Finding: The default agent runtime uses Docker container isolation on GitHub-hosted or Linux self-hosted runners, with the Squid proxy as the only egress path. Eligible GitHub-hosted Ubuntu x86_64 KVM runners can instead place the agent in a Cloud Hypervisor microVM (preview).

Workspace isolation: VM or microVM: Yes

Workspace isolation: VM or microVM

Typed value: Yes.

AWF Docker isolation is the default; opt-in Cloud Hypervisor/NVX preview runtimes use KVM microVMs on eligible runners. The compiler does not provision Git worktrees.

  • Container: The default docker runtime runs the agent in a rootless AWF container.
  • Vm: Preview cloud-hypervisor (GitHub-hosted Ubuntu x86_64 KVM runners) and nvx runtimes place the agent in a KVM microVM.
  • Network: AWF network isolation sends egress only through the proxy to configured domains.
  • MicroVM runtimes are previews limited to eligible KVM runners; the default is a Docker container.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/sandbox.md

    Finding: The default docker runtime runs the agent in a rootless AWF container with network isolation; preview cloud-hypervisor and nvx runtimes use KVM microVMs. Filesystem mounts and outbound network allowlists are configurable. The page does not itself describe the GitHub runner as a VM.

  2. gh-aw: Security architecture

    Finding: The default agent runtime uses Docker container isolation on GitHub-hosted or Linux self-hosted runners, with the Squid proxy as the only egress path. Eligible GitHub-hosted Ubuntu x86_64 KVM runners can instead place the agent in a Cloud Hypervisor microVM (preview).

Parallel runs: Yes cross-checked

Parallel runs

Typed value: Yes.

Workflows on distinct issues/PRs/branches may run concurrently; default per-engine groups serialize each engine.

  • Concurrent sessions: Workflows on distinct issues/PRs/branches may run concurrently; default per-engine groups serialize each engine.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/concurrency.md

    Finding: Separate workflows can run concurrently, with default one agent job per engine. Context-specific groups, cancellation and queue:max preserve or cancel pending runs according to trigger configuration.

Coordinator: Yes

Coordinator

Typed value: Yes.

The product supplies producer/dispatcher agent roles and native queue tools for submitting tasks, dispatching approved worker workflows and inspecting task/claim/result snapshots.

  • Roles: Product-owned producer/dispatcher and worker roles have distinct tools and authority.
  • Results return: Verified Results and Claims are retained in the authoritative Git-backed queue and visible in activation snapshots.
  • Orchestration owner: Native work-queue submission/dispatch and verified worker routes are product orchestration rather than provider-native subagents.
  • Invocation: Producer/dispatcher agents use work_queue_submit and work_queue_dispatch_next; read/explain inspect their immutable activation snapshot.
  • Source review and pinned implementation inspection only; no installed-product, account or multi-user scenario was executed. Unanswered atomic questions remain open.
  1. gh-aw: docs/src/content/docs/reference/work-queue.md

    Finding: Native producer/dispatcher agent roles can submit immutable tasks/dependency graphs and request scheduler-selected assignments to approved worker workflows. work_queue_read/explain inspect the activation snapshot including task, claim and result state. Worker executions are verified/pinned; dependencies consume independently verified Results. Agent tools stage intents; application settlement performs dispatch.

Native issue tracking: Yes cross-checked

Native issue tracking

Typed value: Yes.

The native work queue (tools.work-queue) keeps Work items with dependencies, priorities and attempt state in a Git-backed ledger, separate from individual workflow runs. Operators submit, list, filter, search, dispatch, cancel and reprioritize Work with gh aw work-queue; GitHub Issues can mirror items but are not the store.

  • Create: submit-work and submit-graph admit Work items, optionally as a dependency graph.
  • Prioritize: Work has a priority (default 3); administrators can reprioritize Work from 1 to 5.
  • Search: state --search TEXT searches the queue tree.
  • Filter: state filters by graph, pool and state.
  • Explicit dispatch: dispatch-next commits assignments and dispatch activates a verified dispatcher workflow for queued Work.
  • Task session separation: Work is an immutable task definition; each attempt is a separate Claim bound to a workflow run.
  • Durable states: The version-3 transaction log on the queue branch is authoritative and records Claims, Completions and Results.
  • External tracker requirement: No external tracker is required: Issues can mirror Work but cannot store the queue.
  • Discussion, labels, assignment to humans, archive and reopen were not established for queue Work. The queue is agent-worker infrastructure rather than a human backlog UI. No installed-product test was run.
  1. GitHub Agentic Workflows reference: Work queues (v1.2.2)

    Finding: tools.work-queue is a native Git-backed queue whose transaction log is authoritative; it stores Work task definitions with dependencies, Claims for each attempt, Completions and Results. GitHub Issues can mirror tasks but cannot store the queue.

  2. GitHub Agentic Workflows reference: Work queue operator commands (v1.2.2)

    Finding: gh aw work-queue offers operator commands to submit Work, list state filtered by graph, pool, state or search text, inspect items, dispatch assignments, cancel Work or Claims and, for administrators, reprioritize Work; a keyboard TUI browses the queue.

Unattended triggers: Yes cross-checked

Unattended triggers

Typed value: Yes.

Schedule frontmatter can run a recurring workflow such as a daily report. GitHub events can start automation without a human mentioning the agent on every run. Schedules execute through GitHub Actions rather than a local browser or desktop scheduler.

  • Schedules: Schedule frontmatter can run a recurring workflow such as a daily report.
  • Events: GitHub events can start automation without a human mentioning the agent on every run.
  • Scheduler location: Schedules execute through GitHub Actions rather than a local browser or desktop scheduler.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/triggers.md

    Finding: Workflow frontmatter supports scheduled and event-driven repository automation.

Shared prompting: No

Shared prompting

Typed value: No.

The scoped workflow runtime accepts its invocation inputs and does not supply multiple humans with a shared live prompting channel. GitHub discussion and follow-on workflows are external platform interactions.

  • Coprompting: No product shared live-prompt channel to an existing agent run in the scoped extension/runtime.
  • Source review and pinned implementation inspection only; no installed-product, account or multi-user scenario was executed. Unanswered atomic questions remain open.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Team roles: No

Team roles

Typed value: No.

Human authentication and repository authorization belong to GitHub. Native queue observer/producer/worker roles are agent authority, not human workspace roles or product SSO.

  • Organization roles: No separate gh-aw human workspace/team-role model.
  • Sso: No product SSO implementation; GitHub login and enterprise authentication are external platform dependencies.
  • Source review and pinned implementation inspection only; no installed-product, account or multi-user scenario was executed. Unanswered atomic questions remain open.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

SSO: No

SSO

Typed value: No.

Human authentication and repository authorization belong to GitHub. Native queue observer/producer/worker roles are agent authority, not human workspace roles or product SSO.

  • Organization roles: No separate gh-aw human workspace/team-role model.
  • Sso: No product SSO implementation; GitHub login and enterprise authentication are external platform dependencies.
  • Source review and pinned implementation inspection only; no installed-product, account or multi-user scenario was executed. Unanswered atomic questions remain open.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Pull request workflow: Yes cross-checked

Pull request workflow

Typed value: Yes.

create-pull-request safe outputs propose code in a separate permission-scoped job. Workflow examples include CI-failure investigation; automatically triggering CI from generated PRs may require an additional token.

  • Pull requests: create-pull-request safe outputs propose code in a separate permission-scoped job.
  • Ci response: Workflow examples include CI-failure investigation; automatically triggering CI from generated PRs may require an additional token.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/safe-outputs-pull-requests.md

    Finding: Permission-scoped safe outputs can create and manage pull requests. Policy, credentials and individual output configuration govern each operation.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Review interface: No

Review interface

Typed value: No.

The extension exposes workflow logs, audit reports and output artifacts, not a product UI for reviewing proposed code diffs. GitHub pull-request review is separate.

  • Diffs: No product-owned proposed-diff review UI in the scoped extension/runtime; reports and GitHub PR pages are separate.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Product license: Apache-2.0: No. MIT cross-checked

Product license: Apache-2.0

Typed value: No. MIT.

The gh-aw compiler and runtime repository use MIT.

  • Client license: The gh-aw compiler and runtime repository use MIT.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows LICENSE

    Finding: The gh-aw repository code is MIT licensed.

Product license: MIT: Yes. MIT cross-checked

Product license: MIT

Typed value: Yes. MIT.

The gh-aw compiler and runtime repository use MIT.

  • Client license: The gh-aw compiler and runtime repository use MIT.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows LICENSE

    Finding: The gh-aw repository code is MIT licensed.

Product license: GPL-3.0: No. MIT cross-checked

Product license: GPL-3.0

Typed value: No. MIT.

The gh-aw compiler and runtime repository use MIT.

  • Client license: The gh-aw compiler and runtime repository use MIT.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows LICENSE

    Finding: The gh-aw repository code is MIT licensed.

Product license: GPL-3.0-or-later: No. MIT cross-checked

Product license: GPL-3.0-or-later

Typed value: No. MIT.

The gh-aw compiler and runtime repository use MIT.

  • Client license: The gh-aw compiler and runtime repository use MIT.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows LICENSE

    Finding: The gh-aw repository code is MIT licensed.

Product license: AGPL-3.0: No. MIT cross-checked

Product license: AGPL-3.0

Typed value: No. MIT.

The gh-aw compiler and runtime repository use MIT.

  • Client license: The gh-aw compiler and runtime repository use MIT.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows LICENSE

    Finding: The gh-aw repository code is MIT licensed.

Product license: AGPL-3.0-or-later: No. MIT cross-checked

Product license: AGPL-3.0-or-later

Typed value: No. MIT.

The gh-aw compiler and runtime repository use MIT.

  • Client license: The gh-aw compiler and runtime repository use MIT.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows LICENSE

    Finding: The gh-aw repository code is MIT licensed.

Product license: Elastic License 2.0: No. MIT cross-checked

Product license: Elastic License 2.0

Typed value: No. MIT.

The gh-aw compiler and runtime repository use MIT.

  • Client license: The gh-aw compiler and runtime repository use MIT.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows LICENSE

    Finding: The gh-aw repository code is MIT licensed.

Product license: Custom terms: No. MIT cross-checked

Product license: Custom terms

Typed value: No. MIT.

The gh-aw compiler and runtime repository use MIT.

  • Client license: The gh-aw compiler and runtime repository use MIT.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows LICENSE

    Finding: The gh-aw repository code is MIT licensed.

Product license: Proprietary: No. MIT cross-checked

Product license: Proprietary

Typed value: No. MIT.

The gh-aw compiler and runtime repository use MIT.

  • Client license: The gh-aw compiler and runtime repository use MIT.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows LICENSE

    Finding: The gh-aw repository code is MIT licensed.

Telemetry opt-out: Yes

Telemetry opt-out

Typed value: Yes.

gh-aw workflow OTLP telemetry export is configurable and can be disabled by removing workflow/import endpoints and clearing default endpoint secrets/variables. GitHub platform and coding-engine data policies remain separate.

  • Telemetry controls: Clear both default endpoint values and remove workflow/import observability.otlp; export becomes a no-op.
  • Provider separation: The control disables gh-aw OTLP export, not GitHub operational logs or coding-agent telemetry.
  • Source review and pinned implementation inspection only; no installed-product, account or multi-user scenario was executed. Unanswered atomic questions remain open.
  1. gh-aw: docs/src/content/docs/reference/open-telemetry.mdx

    Finding: Removing workflow/import observability.otlp and clearing both the secret and variable GH_AW_DEFAULT_OTLP_ENDPOINT disables gh-aw OTLP export. With both endpoint values unset export resolves to an empty endpoint and is a no-op. GitHub platform logs and coding-engine telemetry are separate.

Free access: Yes cross-checked

Free access

Typed value: Yes.

gh-aw itself is free; model and runner charges remain separate.

  • Ongoing free: gh-aw itself is free; model and runner charges remain separate.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Starting price: Free: Yes. Free cross-checked

Starting price: Free

Typed value: Yes. Free.

gh-aw itself is free; model and runner charges remain separate.

  • Ongoing free: gh-aw itself is free; model and runner charges remain separate.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Starting price: Up to $20/month: No. Free cross-checked

Starting price: Up to $20/month

Typed value: No. Free.

gh-aw itself is free; model and runner charges remain separate.

  • Ongoing free: gh-aw itself is free; model and runner charges remain separate.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Starting price: $21–40/month: No. Free cross-checked

Starting price: $21–40/month

Typed value: No. Free.

gh-aw itself is free; model and runner charges remain separate.

  • Ongoing free: gh-aw itself is free; model and runner charges remain separate.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Starting price: $41–100/month: No. Free cross-checked

Starting price: $41–100/month

Typed value: No. Free.

gh-aw itself is free; model and runner charges remain separate.

  • Ongoing free: gh-aw itself is free; model and runner charges remain separate.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Starting price: Over $100/month: No. Free cross-checked

Starting price: Over $100/month

Typed value: No. Free.

gh-aw itself is free; model and runner charges remain separate.

  • Ongoing free: gh-aw itself is free; model and runner charges remain separate.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Starting price: Contact for pricing: No. Free cross-checked

Starting price: Contact for pricing

Typed value: No. Free.

gh-aw itself is free; model and runner charges remain separate.

  • Ongoing free: gh-aw itself is free; model and runner charges remain separate.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Pricing model: Free, no paid plan: Yes

Pricing model: Free, no paid plan

Typed value: Yes.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Pricing model: Free plan: No

Pricing model: Free plan

Typed value: No.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Workspace fees: No fee for the scoped MIT extension/compiler software; platform runner and inference charges are separate.
  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

  3. gh-aw MIT license

    Finding: The main extension/compiler software is MIT licensed and explicitly permits use free of charge; Actions runner compute and inference plans are separate dependencies.

Pricing model: Subscription: No

Pricing model: Subscription

Typed value: No.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Workspace fees: No fee for the scoped MIT extension/compiler software; platform runner and inference charges are separate.
  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

  3. gh-aw MIT license

    Finding: The main extension/compiler software is MIT licensed and explicitly permits use free of charge; Actions runner compute and inference plans are separate dependencies.

Pricing model: Per seat: No

Pricing model: Per seat

Typed value: No.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Workspace fees: No fee for the scoped MIT extension/compiler software; platform runner and inference charges are separate.
  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

  3. gh-aw MIT license

    Finding: The main extension/compiler software is MIT licensed and explicitly permits use free of charge; Actions runner compute and inference plans are separate dependencies.

Pricing model: Usage-based: No

Pricing model: Usage-based

Typed value: No.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Workspace fees: No fee for the scoped MIT extension/compiler software; platform runner and inference charges are separate.
  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

  3. gh-aw MIT license

    Finding: The main extension/compiler software is MIT licensed and explicitly permits use free of charge; Actions runner compute and inference plans are separate dependencies.

Pricing model: Per machine or compute: No

Pricing model: Per machine or compute

Typed value: No.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Workspace fees: No fee for the scoped MIT extension/compiler software; platform runner and inference charges are separate.
  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

  3. gh-aw MIT license

    Finding: The main extension/compiler software is MIT licensed and explicitly permits use free of charge; Actions runner compute and inference plans are separate dependencies.

Pricing model: Bring your own key: Yes cross-checked

Pricing model: Bring your own key

Typed value: Yes.

A personal Copilot subscription can use COPILOT_GITHUB_TOKEN. Claude subscription OAuth and ChatGPT subscription login are unsupported. Claude and Codex support API credentials and provider-specific inference routes.

  • Agent subscriptions: A personal Copilot subscription can use COPILOT_GITHUB_TOKEN. Claude subscription OAuth and ChatGPT subscription login are unsupported.
  • Api accounts: Claude and Codex support API credentials and provider-specific inference routes.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/engines/claude.md

    Finding: Claude supports API/WIF or Copilot inference, not Claude subscription OAuth. dontAsk is default, tools.edit false denies edits, and MCP gateway permissions remain separate from agent approval modes.

  2. GitHub Agentic Workflows docs/src/content/docs/engines/codex.md

    Finding: Codex uses OpenAI API credentials or Copilot inference; ChatGPT subscription login is unsupported. Per-command Bash allowlists are unavailable for Codex.

Pricing model: One-time purchase: No

Pricing model: One-time purchase

Typed value: No.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Workspace fees: No fee for the scoped MIT extension/compiler software; platform runner and inference charges are separate.
  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

  3. gh-aw MIT license

    Finding: The main extension/compiler software is MIT licensed and explicitly permits use free of charge; Actions runner compute and inference plans are separate dependencies.

Paid plans from: No paid plan: Yes. No paid plan

Paid plans from: No paid plan

Typed value: Yes. No paid plan.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Paid plans from: $10 or less/month: No. No paid plan

Paid plans from: $10 or less/month

Typed value: No. No paid plan.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Paid plans from: $11–20/month: No. No paid plan

Paid plans from: $11–20/month

Typed value: No. No paid plan.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Paid plans from: $21–40/month: No. No paid plan

Paid plans from: $21–40/month

Typed value: No. No paid plan.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Paid plans from: $41–100/month: No. No paid plan

Paid plans from: $41–100/month

Typed value: No. No paid plan.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Paid plans from: Over $100/month: No. No paid plan

Paid plans from: Over $100/month

Typed value: No. No paid plan.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Paid plans from: Contact sales: No. No paid plan

Paid plans from: Contact sales

Typed value: No. No paid plan.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Keeps going with laptop closed: Yes

Keeps going with laptop closed

Typed value: Yes.

Compiled workflows execute on GitHub Actions runners, including user-controlled Linux runners. A runner on the user computer executes there; other self-hosted runners and GitHub-hosted runners remain separate options.

  • Client disconnect: Event and scheduled workflows execute on the Actions runner after the initiating client disconnects.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows README.md

    Finding: The gh aw extension compiles Markdown/YAML repository automation into GitHub Actions workflows with configurable agents, read-only sandbox defaults and separate safe-output jobs.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/triggers.md

    Finding: Workflow frontmatter supports scheduled and event-driven repository automation.

Open-source license: Yes. MIT cross-checked

Open-source license

Typed value: Yes. MIT.

The gh-aw compiler and runtime repository use MIT.

  • Client license: The gh-aw compiler and runtime repository use MIT.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows LICENSE

    Finding: The gh-aw repository code is MIT licensed.

Any desktop app: Yes

Any desktop app

Typed value: Yes.

gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

  • Desktop os: Native release extension binaries cover macOS, Windows and Linux; Windows installer uses Git Bash/MSYS/Cygwin.
  • Mobile apps: No native iOS/Android application; Android Termux CLI installation is not a native phone app.
  • Browser: GitHub workflow pages are the external platform UI; the extension does not supply an ADE browser client.
  • Mobile browser: No product-owned phone-browser ADE interface.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

Paid plan up to $20/month: No. No paid plan

Paid plan up to $20/month

Typed value: No. No paid plan.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Paid plan up to $40/month: No. No paid plan

Paid plan up to $40/month

Typed value: No. No paid plan.

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • Model cost: Inference is charged by the selected provider or Copilot plan, separate from Actions minutes.
  • Compute cost: GitHub Actions compute minutes are separate from model inference.
  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

Any phone client: No

Any phone client

Typed value: No.

gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

  • Desktop os: Native release extension binaries cover macOS, Windows and Linux; Windows installer uses Git Bash/MSYS/Cygwin.
  • Mobile apps: No native iOS/Android application; Android Termux CLI installation is not a native phone app.
  • Browser: GitHub workflow pages are the external platform UI; the extension does not supply an ADE browser client.
  • Mobile browser: No product-owned phone-browser ADE interface.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

All public findings

Findings may summarize a source. They are not presented as verbatim statements.

gh-aw:ACC-01: gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

gh-aw:ACC-01

gh aw ships a native GitHub CLI extension for macOS, Windows and Linux; this is CLI/TUI access, not a separate desktop GUI. It supplies no released product browser or native phone app.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows README.md

    Finding: The gh aw extension compiles Markdown/YAML repository automation into GitHub Actions workflows with configurable agents, read-only sandbox defaults and separate safe-output jobs.

  2. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

gh-aw:ACC-02: Built-in engines and custom imported definitions are separate. Public experimental samples launch Cursor Agent, OpenCode and Aider without compatibility guarantees; these are runnable integrations, not supported built-in engines.

gh-aw:ACC-02

Built-in engines and custom imported definitions are separate. Public experimental samples launch Cursor Agent, OpenCode and Aider without compatibility guarantees; these are runnable integrations, not supported built-in engines.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  • The released Agy engine is experimental, Linux x64 only, and supports Gemini API keys rather than Google WIF, ADC or Vertex AI. It cannot disable native shell, editing or web tools.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/engines.md

    Finding: Built-in engines include Copilot, Claude Code, Codex, deprecated Gemini CLI and Pi. Experimental durable mode checkpoints native sessions in artifacts; max-turns and AI-credit limits vary by engine.

  2. gh-aw Antigravity support at v1.2.2

    Finding: gh-aw v1.2.2 ships the native experimental Agy engine and documents engine: agy running Antigravity CLI in GitHub Actions.

  3. gh-aw Antigravity support at v1.2.2

    Finding: gh-aw v1.2.2 ships the native experimental Agy engine and documents engine: agy running Antigravity CLI in GitHub Actions.

  4. gh-aw: .github/workflows/shared/cursor.md

    Finding: The repository ships publicly importable experimental engine definitions that install/launch Cursor Agent, OpenCode 2 and Aider. They are explicitly unsupported examples for integration authors without a compatibility or maintenance commitment, rather than built-in supported engines. Custom engine definitions are accepted by the compiler.

gh-aw:ACC-03: A personal Copilot subscription can use COPILOT_GITHUB_TOKEN. Claude subscription OAuth and ChatGPT subscription login are unsupported. Claude and Codex support API credentials and provider-specific inference routes. cross-checked

gh-aw:ACC-03

A personal Copilot subscription can use COPILOT_GITHUB_TOKEN. Claude subscription OAuth and ChatGPT subscription login are unsupported. Claude and Codex support API credentials and provider-specific inference routes.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/engines/claude.md

    Finding: Claude supports API/WIF or Copilot inference, not Claude subscription OAuth. dontAsk is default, tools.edit false denies edits, and MCP gateway permissions remain separate from agent approval modes.

  2. GitHub Agentic Workflows docs/src/content/docs/engines/codex.md

    Finding: Codex uses OpenAI API credentials or Copilot inference; ChatGPT subscription login is unsupported. Per-command Bash allowlists are unavailable for Codex.

  3. GitHub Agentic Workflows docs/src/content/docs/engines/copilot.md

    Finding: Personal Copilot subscriptions can be used with COPILOT_GITHUB_TOKEN, separate from organization-billed inference.

gh-aw:ACC-04: Install the extension, select/authenticate an engine, add Markdown workflow source and compile its lock file. The generated workflow runs on GitHub Actions against the configured repository. cross-checked

gh-aw:ACC-04

Install the extension, select/authenticate an engine, add Markdown workflow source and compile its lock file. The generated workflow runs on GitHub Actions against the configured repository.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows README.md

    Finding: The gh aw extension compiles Markdown/YAML repository automation into GitHub Actions workflows with configurable agents, read-only sandbox defaults and separate safe-output jobs.

gh-aw:ACC-05: gh aw compile, logs and audit provide workflow authoring and inspection operations. gh aw mcp-server exposes workflow operations, read-only unless write tools are enabled. cross-checked

gh-aw:ACC-05

gh aw compile, logs and audit provide workflow authoring and inspection operations. gh aw mcp-server exposes workflow operations, read-only unless write tools are enabled.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows README.md

    Finding: The gh aw extension compiles Markdown/YAML repository automation into GitHub Actions workflows with configurable agents, read-only sandbox defaults and separate safe-output jobs.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  3. GitHub Agentic Workflows docs/src/content/docs/reference/gh-aw-as-mcp-server.md

    Finding: gh aw mcp-server exposes workflow-management CLI operations as MCP tools, read-only by default; writes require --allow-write-tools.

gh-aw:ACC-06: Declared MCP servers are filtered through the configured gateway tool allowlist. Engine-specific plugins are supported where listed; agent plugins remain experimental. Gateway allowlists constrain MCP calls independently of Claude permission mode. cross-checked

gh-aw:ACC-06

Declared MCP servers are filtered through the configured gateway tool allowlist. Engine-specific plugins are supported where listed; agent plugins remain experimental. Gateway allowlists constrain MCP calls independently of Claude permission mode.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/engines.md

    Finding: Built-in engines include Copilot, Claude Code, Codex, deprecated Gemini CLI and Pi. Experimental durable mode checkpoints native sessions in artifacts; max-turns and AI-credit limits vary by engine.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/engines/claude.md

    Finding: Claude supports API/WIF or Copilot inference, not Claude subscription OAuth. dontAsk is default, tools.edit false denies edits, and MCP gateway permissions remain separate from agent approval modes.

gh-aw:PLAN-01: Workflow Markdown holds natural-language task instructions; YAML frontmatter sets triggers and policy. Issues, PRs, comments and other configured GitHub events can trigger workflows. cross-checked

gh-aw:PLAN-01

Workflow Markdown holds natural-language task instructions; YAML frontmatter sets triggers and policy. Issues, PRs, comments and other configured GitHub events can trigger workflows.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows README.md

    Finding: The gh aw extension compiles Markdown/YAML repository automation into GitHub Actions workflows with configurable agents, read-only sandbox defaults and separate safe-output jobs.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/triggers.md

    Finding: Workflow frontmatter supports scheduled and event-driven repository automation.

gh-aw:PLAN-04: Cache-memory persists declared files between runs with workflow scope/retention. cross-checked

gh-aw:PLAN-04

Cache-memory persists declared files between runs with workflow scope/retention.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cache-memory.md

    Finding: Workflow cache-memory persists declared files across workflow runs with configured scope and retention.

gh-aw:PLAN-06: The native work queue (tools.work-queue) keeps Work items with dependencies, priorities and attempt state in a Git-backed ledger, separate from individual workflow runs. Operators submit, list, filter, search, dispatch, cancel and reprioritize Work with gh aw work-queue; GitHub Issues can mirror items but are not the store. cross-checked

gh-aw:PLAN-06

The native work queue (tools.work-queue) keeps Work items with dependencies, priorities and attempt state in a Git-backed ledger, separate from individual workflow runs. Operators submit, list, filter, search, dispatch, cancel and reprioritize Work with gh aw work-queue; GitHub Issues can mirror items but are not the store.

  • Discussion, labels, assignment to humans, archive and reopen were not established for queue Work. The queue is agent-worker infrastructure rather than a human backlog UI. No installed-product test was run.
  1. GitHub Agentic Workflows reference: Work queues (v1.2.2)

    Finding: tools.work-queue is a native Git-backed queue whose transaction log is authoritative; it stores Work task definitions with dependencies, Claims for each attempt, Completions and Results. GitHub Issues can mirror tasks but cannot store the queue.

  2. GitHub Agentic Workflows reference: Work queue operator commands (v1.2.2)

    Finding: gh aw work-queue offers operator commands to submit Work, list state filtered by graph, pool, state or search text, inspect items, dispatch assignments, cancel Work or Claims and, for administrators, reprioritize Work; a keyboard TUI browses the queue.

gh-aw:COORD-01: Workflows on distinct issues/PRs/branches may run concurrently; default per-engine groups serialize each engine. cross-checked

gh-aw:COORD-01

Workflows on distinct issues/PRs/branches may run concurrently; default per-engine groups serialize each engine.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/concurrency.md

    Finding: Separate workflows can run concurrently, with default one agent job per engine. Context-specific groups, cancellation and queue:max preserve or cancel pending runs according to trigger configuration.

gh-aw:COORD-02: queue:max and cancel-in-progress settings determine pending-run retention and obsolete-PR cancellation. cross-checked

gh-aw:COORD-02

queue:max and cancel-in-progress settings determine pending-run retention and obsolete-PR cancellation.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/concurrency.md

    Finding: Separate workflows can run concurrently, with default one agent job per engine. Context-specific groups, cancellation and queue:max preserve or cancel pending runs according to trigger configuration.

gh-aw:COORD-03: The product supplies producer/dispatcher agent roles and native queue tools for submitting tasks, dispatching approved worker workflows and inspecting task/claim/result snapshots.

gh-aw:COORD-03

The product supplies producer/dispatcher agent roles and native queue tools for submitting tasks, dispatching approved worker workflows and inspecting task/claim/result snapshots.

  • Source review and pinned implementation inspection only; no installed-product, account or multi-user scenario was executed. Unanswered atomic questions remain open.
  1. gh-aw: docs/src/content/docs/reference/work-queue.md

    Finding: Native producer/dispatcher agent roles can submit immutable tasks/dependency graphs and request scheduler-selected assignments to approved worker workflows. work_queue_read/explain inspect the activation snapshot including task, claim and result state. Worker executions are verified/pinned; dependencies consume independently verified Results. Agent tools stage intents; application settlement performs dispatch.

gh-aw:COORD-04: Schedule frontmatter can run a recurring workflow such as a daily report. GitHub events can start automation without a human mentioning the agent on every run. Schedules execute through GitHub Actions rather than a local browser or desktop scheduler. cross-checked

gh-aw:COORD-04

Schedule frontmatter can run a recurring workflow such as a daily report. GitHub events can start automation without a human mentioning the agent on every run. Schedules execute through GitHub Actions rather than a local browser or desktop scheduler.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows README.md

    Finding: The gh aw extension compiles Markdown/YAML repository automation into GitHub Actions workflows with configurable agents, read-only sandbox defaults and separate safe-output jobs.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/triggers.md

    Finding: Workflow frontmatter supports scheduled and event-driven repository automation.

gh-aw:RUN-01: Compiled workflows execute on GitHub Actions runners, including user-controlled Linux runners. A runner on the user computer executes there; other self-hosted runners and GitHub-hosted runners remain separate options.

gh-aw:RUN-01

Compiled workflows execute on GitHub Actions runners, including user-controlled Linux runners. A runner on the user computer executes there; other self-hosted runners and GitHub-hosted runners remain separate options.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows README.md

    Finding: The gh aw extension compiles Markdown/YAML repository automation into GitHub Actions workflows with configurable agents, read-only sandbox defaults and separate safe-output jobs.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/sandbox.md

    Finding: The default docker runtime runs the agent in a rootless AWF container with network isolation; preview cloud-hypervisor and nvx runtimes use KVM microVMs. Filesystem mounts and outbound network allowlists are configurable. The page does not itself describe the GitHub runner as a VM.

  3. GitHub Agentic Workflows docs/src/content/docs/reference/triggers.md

    Finding: Workflow frontmatter supports scheduled and event-driven repository automation.

  4. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

  5. gh-aw: Security architecture

    Finding: The default agent runtime uses Docker container isolation on GitHub-hosted or Linux self-hosted runners, with the Squid proxy as the only egress path. Eligible GitHub-hosted Ubuntu x86_64 KVM runners can instead place the agent in a Cloud Hypervisor microVM (preview).

  6. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

gh-aw:RUN-02: AWF Docker isolation is the default; opt-in Cloud Hypervisor/NVX preview runtimes use KVM microVMs on eligible runners. The compiler does not provision Git worktrees.

gh-aw:RUN-02

AWF Docker isolation is the default; opt-in Cloud Hypervisor/NVX preview runtimes use KVM microVMs on eligible runners. The compiler does not provision Git worktrees.

  • MicroVM runtimes are previews limited to eligible KVM runners; the default is a Docker container.
  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/sandbox.md

    Finding: The default docker runtime runs the agent in a rootless AWF container with network isolation; preview cloud-hypervisor and nvx runtimes use KVM microVMs. Filesystem mounts and outbound network allowlists are configurable. The page does not itself describe the GitHub runner as a VM.

  2. gh-aw: Security architecture

    Finding: The default agent runtime uses Docker container isolation on GitHub-hosted or Linux self-hosted runners, with the Squid proxy as the only egress path. Eligible GitHub-hosted Ubuntu x86_64 KVM runners can instead place the agent in a Cloud Hypervisor microVM (preview).

  3. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

gh-aw:RUN-03: Experimental engine.durable saves/restores native session checkpoints in artifacts, default seven days subject to repository retention. Checkpoint format and resume behavior differ by engine; durable mode does not automatically dispatch a continuation. cross-checked

gh-aw:RUN-03

Experimental engine.durable saves/restores native session checkpoints in artifacts, default seven days subject to repository retention. Checkpoint format and resume behavior differ by engine; durable mode does not automatically dispatch a continuation.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/engines.md

    Finding: Built-in engines include Copilot, Claude Code, Codex, deprecated Gemini CLI and Pi. Experimental durable mode checkpoints native sessions in artifacts; max-turns and AI-credit limits vary by engine.

gh-aw:RUN-05: Top-level max-turns defaults to 500 for supported engines; Agy lacks this limit. max-ai-credits bounds catalog-estimated inference consumption through the proxy. Claude recovery resumes the exact session; silent fresh restart after progress requires an explicit override. cross-checked

gh-aw:RUN-05

Top-level max-turns defaults to 500 for supported engines; Agy lacks this limit. max-ai-credits bounds catalog-estimated inference consumption through the proxy. Claude recovery resumes the exact session; silent fresh restart after progress requires an explicit override.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/engines.md

    Finding: Built-in engines include Copilot, Claude Code, Codex, deprecated Gemini CLI and Pi. Experimental durable mode checkpoints native sessions in artifacts; max-turns and AI-credit limits vary by engine.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/engines/claude.md

    Finding: Claude supports API/WIF or Copilot inference, not Claude subscription OAuth. dontAsk is default, tools.edit false denies edits, and MCP gateway permissions remain separate from agent approval modes.

  3. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

gh-aw:RUN-06: gh aw logs and audit expose per-run duration, token/credit usage and tool/session information. cross-checked

gh-aw:RUN-06

gh aw logs and audit expose per-run duration, token/credit usage and tool/session information.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

gh-aw:RUN-07: AI Credits estimate inference per workflow run; Actions compute is a separate cost component. Credit estimates use catalog prices and do not reconcile the providers invoice. Structured logs JSON includes per-run aic and episode total_aic fields. cross-checked

gh-aw:RUN-07

AI Credits estimate inference per workflow run; Actions compute is a separate cost component. Credit estimates use catalog prices and do not reconcile the providers invoice. Structured logs JSON includes per-run aic and episode total_aic fields.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

gh-aw:RUN-08: Workflow and engine concurrency groups bound overlap and preserve queued work where configured. PR updates can cancel outdated runs; mixed trigger queue behavior is documented separately. cross-checked

gh-aw:RUN-08

Workflow and engine concurrency groups bound overlap and preserve queued work where configured. PR updates can cancel outdated runs; mixed trigger queue behavior is documented separately.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/concurrency.md

    Finding: Separate workflows can run concurrently, with default one agent job per engine. Context-specific groups, cancellation and queue:max preserve or cancel pending runs according to trigger configuration.

gh-aw:TEAM-01: The scoped workflow runtime accepts its invocation inputs and does not supply multiple humans with a shared live prompting channel. GitHub discussion and follow-on workflows are external platform interactions.

gh-aw:TEAM-01

The scoped workflow runtime accepts its invocation inputs and does not supply multiple humans with a shared live prompting channel. GitHub discussion and follow-on workflows are external platform interactions.

  • Source review and pinned implementation inspection only; no installed-product, account or multi-user scenario was executed. Unanswered atomic questions remain open.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

gh-aw:TEAM-02: Human authentication and repository authorization belong to GitHub. Native queue observer/producer/worker roles are agent authority, not human workspace roles or product SSO.

gh-aw:TEAM-02

Human authentication and repository authorization belong to GitHub. Native queue observer/producer/worker roles are agent authority, not human workspace roles or product SSO.

  • Source review and pinned implementation inspection only; no installed-product, account or multi-user scenario was executed. Unanswered atomic questions remain open.
  1. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

gh-aw:TEAM-03: The gateway applies explicit tool allowlists; engine-level tool enforcement differs by engine. Firewall allowlists and scoped safe-output jobs provide configured isolation/write controls. Agents default to read-only access; GitHub mutations normally occur in separately scoped safe-output jobs.

gh-aw:TEAM-03

The gateway applies explicit tool allowlists; engine-level tool enforcement differs by engine. Firewall allowlists and scoped safe-output jobs provide configured isolation/write controls. Agents default to read-only access; GitHub mutations normally occur in separately scoped safe-output jobs.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows README.md

    Finding: The gh aw extension compiles Markdown/YAML repository automation into GitHub Actions workflows with configurable agents, read-only sandbox defaults and separate safe-output jobs.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/engines/claude.md

    Finding: Claude supports API/WIF or Copilot inference, not Claude subscription OAuth. dontAsk is default, tools.edit false denies edits, and MCP gateway permissions remain separate from agent approval modes.

  3. GitHub Agentic Workflows docs/src/content/docs/engines/codex.md

    Finding: Codex uses OpenAI API credentials or Copilot inference; ChatGPT subscription login is unsupported. Per-command Bash allowlists are unavailable for Codex.

  4. GitHub Agentic Workflows docs/src/content/docs/introduction/architecture.mdx

    Finding: Agent execution, sandbox firewall, tool gateway and validated safe-output jobs form separate trust boundaries. Controls are configurable.

  5. GitHub Agentic Workflows docs/src/content/docs/reference/sandbox.md

    Finding: The default docker runtime runs the agent in a rootless AWF container with network isolation; preview cloud-hypervisor and nvx runtimes use KVM microVMs. Filesystem mounts and outbound network allowlists are configurable. The page does not itself describe the GitHub runner as a VM.

gh-aw:SHIP-01: The extension exposes workflow logs, audit reports and output artifacts, not a product UI for reviewing proposed code diffs. GitHub pull-request review is separate.

gh-aw:SHIP-01

The extension exposes workflow logs, audit reports and output artifacts, not a product UI for reviewing proposed code diffs. GitHub pull-request review is separate.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/engines.md

    Finding: Built-in engines include Copilot, Claude Code, Codex, deprecated Gemini CLI and Pi. Experimental durable mode checkpoints native sessions in artifacts; max-turns and AI-credit limits vary by engine.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  3. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

gh-aw:SHIP-02: Workflow configuration separates deterministic CI jobs from agent reasoning and permitted outputs. cross-checked

gh-aw:SHIP-02

Workflow configuration separates deterministic CI jobs from agent reasoning and permitted outputs.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows README.md

    Finding: The gh aw extension compiles Markdown/YAML repository automation into GitHub Actions workflows with configurable agents, read-only sandbox defaults and separate safe-output jobs.

gh-aw:SHIP-03: create-pull-request safe outputs propose code in a separate permission-scoped job. Workflow examples include CI-failure investigation; automatically triggering CI from generated PRs may require an additional token. cross-checked

gh-aw:SHIP-03

create-pull-request safe outputs propose code in a separate permission-scoped job. Workflow examples include CI-failure investigation; automatically triggering CI from generated PRs may require an additional token.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/safe-outputs-pull-requests.md

    Finding: Permission-scoped safe outputs can create and manage pull requests. Policy, credentials and individual output configuration govern each operation.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

gh-aw:SHIP-04: Merge operations are separately configured safe outputs with scoped credentials; PR creation alone does not grant merging. cross-checked

gh-aw:SHIP-04

Merge operations are separately configured safe outputs with scoped credentials; PR creation alone does not grant merging.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/safe-outputs-pull-requests.md

    Finding: Permission-scoped safe outputs can create and manage pull requests. Policy, credentials and individual output configuration govern each operation.

gh-aw:OWN-01: The gh-aw compiler and runtime repository use MIT. cross-checked

gh-aw:OWN-01

The gh-aw compiler and runtime repository use MIT.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows LICENSE

    Finding: The gh-aw repository code is MIT licensed.

gh-aw:OWN-02: The product code and execution worker can run on user-controlled runners, but GitHub Actions supplies workflow control, triggers, secrets and logs. A self-managed execution worker alone does not establish a fully self-hosted core ADE; a complete user-managed control-plane deployment is unestablished by this source pass.

gh-aw:OWN-02

The product code and execution worker can run on user-controlled runners, but GitHub Actions supplies workflow control, triggers, secrets and logs. A self-managed execution worker alone does not establish a fully self-hosted core ADE; a complete user-managed control-plane deployment is unestablished by this source pass.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

  2. gh-aw: install-gh-aw.sh

    Finding: The install script provides native extension binaries for Linux, macOS, Windows and other hosts. gh aw is a CLI/compiler plus GitHub Actions runtime, not a standalone browser/phone app or diff review interface. It executes agent work on configured Actions runners; a self-hosted runner can be on a user computer. Source checkout is managed by the workflow, with no worktree provisioning. GitHub supplies human accounts, repository permissions, SSO, workflow control and logs; no separate human-team or shared live-prompt service is implemented by the extension.

gh-aw:OWN-03: gh-aw workflow OTLP telemetry export is configurable and can be disabled by removing workflow/import endpoints and clearing default endpoint secrets/variables. GitHub platform and coding-engine data policies remain separate.

gh-aw:OWN-03

gh-aw workflow OTLP telemetry export is configurable and can be disabled by removing workflow/import endpoints and clearing default endpoint secrets/variables. GitHub platform and coding-engine data policies remain separate.

  • Source review and pinned implementation inspection only; no installed-product, account or multi-user scenario was executed. Unanswered atomic questions remain open.
  1. gh-aw: docs/src/content/docs/reference/open-telemetry.mdx

    Finding: Removing workflow/import observability.otlp and clearing both the secret and variable GH_AW_DEFAULT_OTLP_ENDPOINT disables gh-aw OTLP export. With both endpoint values unset export resolves to an empty endpoint and is a no-op. GitHub platform logs and coding-engine telemetry are separate.

gh-aw:OWN-04: The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

gh-aw:OWN-04

The gh-aw extension/compiler is free MIT software without a paid extension plan. Inference is billed by the chosen provider/Copilot plan and Actions compute minutes remain separate.

  • No provider or Actions compute price is inferred from the software license.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/cost-management.md

    Finding: Per-run logs/audit reports display tokens, duration and AI-credit estimates. One AIC represents $0.01 of catalog-estimated inference, distinct from Actions compute and provider invoices.

  2. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

  3. gh-aw MIT license

    Finding: The main extension/compiler software is MIT licensed and explicitly permits use free of charge; Actions runner compute and inference plans are separate dependencies.

gh-aw:OWN-06: Release v1.2.1 dated 2026-11-06 adds experimental artifact-backed continuation and workflow fixes. cross-checked

gh-aw:OWN-06

Release v1.2.1 dated 2026-11-06 adds experimental artifact-backed continuation and workflow fixes.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows v1.2.1

    Finding: Public release v1.2.1 was published 2026-11-06.

gh-aw:OWN-08: gh-aw itself is free; model and runner charges remain separate. cross-checked

gh-aw:OWN-08

gh-aw itself is free; model and runner charges remain separate.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.

gh-aw:OWN-09: The application adds no platform charge; inference and Actions minutes remain separate billed units. cross-checked

gh-aw:OWN-09

The application adds no platform charge; inference and Actions minutes remain separate billed units.

  • Only the cited atomic findings are established. Remaining atomic requirements and the full rubric scenario remain unverified.
  1. GitHub Agentic Workflows docs/src/content/docs/reference/faq.md

    Finding: gh-aw is free with separate inference and Actions compute charges. It requires Linux container-capable runners, can use self-hosted runners, and retains GitHub workflow dependencies.