Agent Deck
Open-source Mac and Linux app. Agents run on your computer or on your own server.
Product websitegithub.com Compare with others Missing or wrong data?
- Desktop app
- Mac and Linux
- Browser
- Yes
- Phone
- Phone browser
- Command line
- Yes
- Agents
- Claude Code, Codex, Cursor Agent, Gemini CLI, OpenCode, Pi and Aider
- Your subscription
- Works with your own plan
- Agents run
- On your computer and on your own server
- Laptop closed
- Keeps going
- Isolation
- Worktree and Container
- Parallel agents
- Yes
- Scheduled runs
- Yes
- Coordinator agent
- Yes
- Built-in issues
- No
- Pull requests
- No
- Review screen
- No
- Team
- Not listed
- Self-hosting
- Yes
- Licence
- MIT
- Price from
- Free
- How it charges
- No paid plan and bring your own key
- Paid plans from
- Not listed
Receipts
- Checked
- · oldest answer 2026-11-08
- Findings
- 35 published · 27 cross-checked by a second agent · read them all
- Sources
- 16 public pages on 1 site: github.com
- Screenshot
- github.com, retrieved · first party repository · shows appearance, not capability
- Data
- /products/agent-deck.json · .md
- Method
- Same rules for all 103 researched ADEs. How we check · Report wrong data
What it does
Where you use it
CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
Source and scope for Where you use itAgents it runs
Many terminal coding agents are supported with different integration levels. At v1.16.38 an aider session command also maps to a built-in Aider tool identity, without status detection.
Source and scope for Agents it runsAccounts and subscriptions
Agent authentication is reused for local/container sessions, with a macOS Claude exception.
Source and scope for Accounts and subscriptionsWhere agents run
Sessions run locally, in Docker or on user-owned SSH/cloud hosts. The scoped execution implementation has no vendor-managed compute service.
Source and scope for Where agents runHow work is kept apart
Git worktrees and Docker containers separate execution scope. The scoped session launcher has no VM or microVM isolation backend.
Source and scope for How work is kept apartParallel work
Independent tmux-backed sessions run concurrently and show process/work state separately.
Source and scope for Parallel workScheduled runs
Heartbeat/watchers and optional chat-service bridges trigger session work.
Source and scope for Scheduled runsGit and pull requests
Product manages worktrees and branches. The scoped application has no built-in PR creation or management command; a coding agent may run its own external commands.
Source and scope for Git and pull requestsPrice
Agent Deck is a completely free MIT application with no paid ADE plan in the checked first-party offering. Underlying coding-agent/model and user-owned compute costs are separate.
Source and scope for PriceTyped filters
Every value links to its receipt below. Read plan and surface restrictions before using a filter to decide.
| Filter | Value |
|---|---|
| Desktop OS: macOS | Yes |
| Desktop OS: Windows | No |
| Desktop OS: Linux | Yes |
| Web client | Yes |
| Phone access: Native iOS | No |
| Phone access: Native Android | No |
| Phone access: Phone browser | Yes |
| CLI | Yes cross-checked |
| Coding agents: Claude Code | Yes cross-checked |
| Coding agents: Codex | Yes cross-checked |
| Coding agents: Cursor Agent | Yes cross-checked |
| Coding agents: Gemini CLI | Yes cross-checked |
| Coding agents: OpenCode | Yes cross-checked |
| Coding agents: Pi | Yes cross-checked |
| Coding agents: Grok | Unknown |
| Coding agents: Aider | Yes cross-checked |
| Coding agents: Antigravity | Unknown |
| Use my subscription | Yes cross-checked |
| Where agents run: On your computer | Yes |
| Where agents run: On your own server | Yes |
| Where agents run: In the vendor's cloud | No |
| Workspace isolation: Git worktree | Yes |
| Workspace isolation: Container | Yes |
| Workspace isolation: VM or microVM | No |
| Parallel runs | Yes cross-checked |
| Coordinator | Yes cross-checked |
| Native issue tracking | No |
| Unattended triggers | Yes cross-checked |
| Shared prompting | Unknown |
| Team roles | No |
| SSO | No |
| Pull request workflow | No |
| Review interface | No |
| Product license: Apache-2.0 | No. MIT cross-checked |
| Product license: MIT | Yes. MIT cross-checked |
| Product license: GPL-3.0 | No. MIT cross-checked |
| Product license: GPL-3.0-or-later | No. MIT cross-checked |
| Product license: AGPL-3.0 | No. MIT cross-checked |
| Product license: AGPL-3.0-or-later | No. MIT cross-checked |
| Product license: Elastic License 2.0 | No. MIT cross-checked |
| Product license: Custom terms | No. MIT cross-checked |
| Product license: Proprietary | No. MIT cross-checked |
| Self-hosting | Yes cross-checked |
| Telemetry opt-out | Yes cross-checked |
| Free access | Yes cross-checked |
| Starting price: Free | Yes. Free cross-checked |
| Starting price: Up to $20/month | No. Free cross-checked |
| Starting price: $21–40/month | No. Free cross-checked |
| Starting price: $41–100/month | No. Free cross-checked |
| Starting price: Over $100/month | No. Free cross-checked |
| Starting price: Contact for pricing | No. Free cross-checked |
| Pricing model: Free, no paid plan | Yes |
| Pricing model: Free plan | No |
| Pricing model: Subscription | No |
| Pricing model: Per seat | No |
| Pricing model: Usage-based | No |
| Pricing model: Per machine or compute | No |
| Pricing model: Bring your own key | Yes |
| Pricing model: One-time purchase | No |
| Paid plans from: No paid plan | Unknown |
| Paid plans from: $10 or less/month | Unknown |
| Paid plans from: $11–20/month | Unknown |
| Paid plans from: $21–40/month | Unknown |
| Paid plans from: $41–100/month | Unknown |
| Paid plans from: Over $100/month | Unknown |
| Paid plans from: Contact sales | Unknown |
| Lifecycle status: Active | Unknown |
| Lifecycle status: Sunsetting | Unknown |
| Keeps going with laptop closed | Yes |
| Open-source license | Yes. MIT cross-checked |
| Any desktop app | Yes |
| Paid plan up to $20/month | Unknown |
| Paid plan up to $40/month | Unknown |
| Any phone client | Yes |
Filter receipts
Desktop OS: macOS: Yes
Desktop OS: macOS
Typed value: Yes.
CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
- Desktop os: macOS and Linux native terminal binaries; Windows requires WSL, with no native Windows binary.
- Mobile apps: No native iOS or Android application in the scoped release build and client tree; phone access uses the browser.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Release builds target darwin and linux on amd64/arm64. The pinned repository release/build tree contains a Go CLI/TUI and embedded browser UI, with no Windows-native, iOS or Android application target. Windows is documented through WSL in README Installation.
Desktop OS: Windows: No
Desktop OS: Windows
Typed value: No.
CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
- Desktop os: macOS and Linux native terminal binaries; Windows requires WSL, with no native Windows binary.
- Mobile apps: No native iOS or Android application in the scoped release build and client tree; phone access uses the browser.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Release builds target darwin and linux on amd64/arm64. The pinned repository release/build tree contains a Go CLI/TUI and embedded browser UI, with no Windows-native, iOS or Android application target. Windows is documented through WSL in README Installation.
Desktop OS: Linux: Yes
Desktop OS: Linux
Typed value: Yes.
CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
- Desktop os: macOS and Linux native terminal binaries; Windows requires WSL, with no native Windows binary.
- Mobile apps: No native iOS or Android application in the scoped release build and client tree; phone access uses the browser.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Release builds target darwin and linux on amd64/arm64. The pinned repository release/build tree contains a Go CLI/TUI and embedded browser UI, with no Windows-native, iOS or Android application target. Windows is documented through WSL in README Installation.
Web client: Yes
Web client
Typed value: Yes.
CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
- Browser: agent-deck web, with optional read-only mode.
- Mobile browser: Home-screen browser app retains authentication cookie; mobile conductor control also uses chat-service bridges.
- Cli: agent-deck terminal UI and subcommands.
- Task actions: Web terminal can interact with sessions; read-only flag restricts input.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Phone access: Native iOS: No
Phone access: Native iOS
Typed value: No.
CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
- Desktop os: macOS and Linux native terminal binaries; Windows requires WSL, with no native Windows binary.
- Mobile apps: No native iOS or Android application in the scoped release build and client tree; phone access uses the browser.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Release builds target darwin and linux on amd64/arm64. The pinned repository release/build tree contains a Go CLI/TUI and embedded browser UI, with no Windows-native, iOS or Android application target. Windows is documented through WSL in README Installation.
Phone access: Native Android: No
Phone access: Native Android
Typed value: No.
CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
- Desktop os: macOS and Linux native terminal binaries; Windows requires WSL, with no native Windows binary.
- Mobile apps: No native iOS or Android application in the scoped release build and client tree; phone access uses the browser.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Release builds target darwin and linux on amd64/arm64. The pinned repository release/build tree contains a Go CLI/TUI and embedded browser UI, with no Windows-native, iOS or Android application target. Windows is documented through WSL in README Installation.
Phone access: Phone browser: Yes
Phone access: Phone browser
Typed value: Yes.
CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
- Browser: agent-deck web, with optional read-only mode.
- Mobile browser: Home-screen browser app retains authentication cookie; mobile conductor control also uses chat-service bridges.
- Cli: agent-deck terminal UI and subcommands.
- Task actions: Web terminal can interact with sessions; read-only flag restricts input.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
CLI: Yes cross-checked
CLI
Typed value: Yes.
CLI exposes session creation, send, stop, fork, parent relationships, MCP, skills, costs and remotes.
- Cli: agent-deck session and related commands.
- Api: Web API and WebSocket terminal routes.
- Operations: Create/start/send/stop/fork/annotate; inspect sessions/remotes.
- Authentication: SSH for remotes; bearer/cookie protection for web.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Coding agents: Claude Code: Yes cross-checked
Coding agents: Claude Code
Typed value: Yes.
Many terminal coding agents are supported with different integration levels. At v1.16.38 an aider session command also maps to a built-in Aider tool identity, without status detection.
- Harnesses: Claude Code, Gemini CLI, Codex, OpenCode 1/2, Cursor, Pi/OMP, Copilot, Crush, Muse and DeepSeek.
- Custom adapters: Custom [tools.*] command configuration.
- Session integration: Launch/status, MCP, fork/resume support differs by tool.
- Mixed harness operation: Several coding CLIs organized in one deck; not hands-on verified.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Coding agents: Codex: Yes cross-checked
Coding agents: Codex
Typed value: Yes.
Many terminal coding agents are supported with different integration levels. At v1.16.38 an aider session command also maps to a built-in Aider tool identity, without status detection.
- Harnesses: Claude Code, Gemini CLI, Codex, OpenCode 1/2, Cursor, Pi/OMP, Copilot, Crush, Muse and DeepSeek.
- Custom adapters: Custom [tools.*] command configuration.
- Session integration: Launch/status, MCP, fork/resume support differs by tool.
- Mixed harness operation: Several coding CLIs organized in one deck; not hands-on verified.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Coding agents: Cursor Agent: Yes cross-checked
Coding agents: Cursor Agent
Typed value: Yes.
Many terminal coding agents are supported with different integration levels. At v1.16.38 an aider session command also maps to a built-in Aider tool identity, without status detection.
- Harnesses: Claude Code, Gemini CLI, Codex, OpenCode 1/2, Cursor, Pi/OMP, Copilot, Crush, Muse and DeepSeek.
- Custom adapters: Custom [tools.*] command configuration.
- Session integration: Launch/status, MCP, fork/resume support differs by tool.
- Mixed harness operation: Several coding CLIs organized in one deck; not hands-on verified.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Coding agents: Gemini CLI: Yes cross-checked
Coding agents: Gemini CLI
Typed value: Yes.
Many terminal coding agents are supported with different integration levels. At v1.16.38 an aider session command also maps to a built-in Aider tool identity, without status detection.
- Harnesses: Claude Code, Gemini CLI, Codex, OpenCode 1/2, Cursor, Pi/OMP, Copilot, Crush, Muse and DeepSeek.
- Custom adapters: Custom [tools.*] command configuration.
- Session integration: Launch/status, MCP, fork/resume support differs by tool.
- Mixed harness operation: Several coding CLIs organized in one deck; not hands-on verified.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Coding agents: OpenCode: Yes cross-checked
Coding agents: OpenCode
Typed value: Yes.
Many terminal coding agents are supported with different integration levels. At v1.16.38 an aider session command also maps to a built-in Aider tool identity, without status detection.
- Harnesses: Claude Code, Gemini CLI, Codex, OpenCode 1/2, Cursor, Pi/OMP, Copilot, Crush, Muse and DeepSeek.
- Custom adapters: Custom [tools.*] command configuration.
- Session integration: Launch/status, MCP, fork/resume support differs by tool.
- Mixed harness operation: Several coding CLIs organized in one deck; not hands-on verified.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Coding agents: Pi: Yes cross-checked
Coding agents: Pi
Typed value: Yes.
Many terminal coding agents are supported with different integration levels. At v1.16.38 an aider session command also maps to a built-in Aider tool identity, without status detection.
- Harnesses: Claude Code, Gemini CLI, Codex, OpenCode 1/2, Cursor, Pi/OMP, Copilot, Crush, Muse and DeepSeek.
- Custom adapters: Custom [tools.*] command configuration.
- Session integration: Launch/status, MCP, fork/resume support differs by tool.
- Mixed harness operation: Several coding CLIs organized in one deck; not hands-on verified.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Coding agents: Aider: Yes cross-checked
Coding agents: Aider
Typed value: Yes.
Many terminal coding agents are supported with different integration levels. At v1.16.38 an aider session command also maps to a built-in Aider tool identity, without status detection.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck session tool mapping
Finding: Creating a session with the command aider resolves to the built-in Aider tool identity and runs the aider command in the session.
-
Agent Deck built-in tool registry
Finding: Aider is a registered built-in tool name with its own icon and colour, but carries no detection patterns, so running aider sessions are not status-detected as Aider.
Use my subscription: Yes cross-checked
Use my subscription
Typed value: Yes.
Agent authentication is reused for local/container sessions, with a macOS Claude exception.
- Agent subscriptions: Container auth shares native Claude/Codex/Gemini sign-ins; macOS Claude uses separate login/setup-token to avoid refresh-chain fork.
- Account switching: Per-group Claude config and per-account agent home directories.
- Vendor account requirement: Core app needs no Agent Deck service account.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Where agents run: On your computer: Yes
Where agents run: On your computer
Typed value: Yes.
Sessions run locally, in Docker or on user-owned SSH/cloud hosts. The scoped execution implementation has no vendor-managed compute service.
- Local: Local tmux sessions.
- Byo remote: remote add/attach/sessions/drain over SSH.
- Self hosted: Own machine runs core app and web UI.
- Other device continuation: Remote host sessions browsed/attached from local deck.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Where agents run: On your own server: Yes
Where agents run: On your own server
Typed value: Yes.
Sessions run locally, in Docker or on user-owned SSH/cloud hosts. The scoped execution implementation has no vendor-managed compute service.
- Local: Local tmux sessions.
- Byo remote: remote add/attach/sessions/drain over SSH.
- Self hosted: Own machine runs core app and web UI.
- Other device continuation: Remote host sessions browsed/attached from local deck.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Where agents run: In the vendor's cloud: No
Where agents run: In the vendor's cloud
Typed value: No.
Sessions run locally, in Docker or on user-owned SSH/cloud hosts. The scoped execution implementation has no vendor-managed compute service.
- Managed cloud: Execution is local or user-managed SSH/Docker; provider connectors use the operator cloud account, without ADE-vendor compute.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck internal/session/instance_sandbox_ssh.go
Finding: Session command preparation runs on the local host, user-configured SSH host or Docker container. The scoped wrapper and container launch code provides no vendor-operated execution service or product VM/microVM isolation backend. Cloud provider connectors use the operator account and remain user-controlled infrastructure.
Workspace isolation: Git worktree: Yes
Workspace isolation: Git worktree
Typed value: Yes.
Git worktrees and Docker containers separate execution scope. The scoped session launcher has no VM or microVM isolation backend.
- Worktree: New worktree per session with setup hooks.
- Container: Run in Docker sandbox option; project bind mounted read-write.
- Secrets: Host tool auth shared to sandbox except separate macOS Claude login.
- Security boundary: Credentials and project remain exposed within declared container mounts; no attack test.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Workspace isolation: Container: Yes
Workspace isolation: Container
Typed value: Yes.
Git worktrees and Docker containers separate execution scope. The scoped session launcher has no VM or microVM isolation backend.
- Worktree: New worktree per session with setup hooks.
- Container: Run in Docker sandbox option; project bind mounted read-write.
- Secrets: Host tool auth shared to sandbox except separate macOS Claude login.
- Security boundary: Credentials and project remain exposed within declared container mounts; no attack test.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Workspace isolation: VM or microVM: No
Workspace isolation: VM or microVM
Typed value: No.
Git worktrees and Docker containers separate execution scope. The scoped session launcher has no VM or microVM isolation backend.
- Vm: The full session command wrapping and sandbox implementation supports Docker, with no product-owned VM/microVM backend.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck internal/session/instance_sandbox_ssh.go
Finding: Session command preparation runs on the local host, user-configured SSH host or Docker container. The scoped wrapper and container launch code provides no vendor-operated execution service or product VM/microVM isolation backend. Cloud provider connectors use the operator account and remain user-controlled infrastructure.
Parallel runs: Yes cross-checked
Parallel runs
Typed value: Yes.
Independent tmux-backed sessions run concurrently and show process/work state separately.
- Concurrent sessions: Multiple coding tools and sessions run in background.
- Status visibility: Running/Waiting/Error/Idle process state plus annotated semantic work state.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Coordinator: Yes cross-checked
Coordinator
Typed value: Yes.
Persistent conductor sessions orchestrate workers and receive status nudges.
- Roles: Named conductors using Claude or Codex and worker sessions.
- Messages: CLI sends prompts; parent nudges on child status transitions.
- Results return: Conductor can inspect sessions and remote drain pulls finished/stalled reports.
- Cross harness: Conductor can manage supported terminal tools.
- Orchestration owner: Agent Deck conductor, notifier and bridge daemon.
- Live progress: Live status and terminal output.
- Followup: Send another prompt or restart session.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Native issue tracking: No
Native issue tracking
Typed value: No.
Fleet board cards and work states are session annotations. The scoped persistence and CLI model has no independent native issue/task entity.
- Task session separation: Instances and groups own work annotations; run-task completes a child session worker, without a separate issue entity.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck internal/statedb/statedb.go
Finding: Persistent state models agent instances, groups, session claims, history, watchers and events. Fleet work states annotate instances; no separate product-owned issue/task entity exists. The run-task command wraps a session worker and records its completion rather than creating an independent issue.
Unattended triggers: Yes cross-checked
Unattended triggers
Typed value: Yes.
Heartbeat/watchers and optional chat-service bridges trigger session work.
- Events: Parent notifications and Telegram/Slack/Discord/Mattermost commands.
- Recurring tasks: Conductor setup installs periodic heartbeat.
- Scheduler location: Bridge/heartbeat/notify daemons on user host.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Team roles: No
Team roles
Typed value: No.
Access uses a shared instance bearer token and global read-only mode. The scoped application has no named human workspace roles or enterprise SSO.
- Organization roles: The complete web authentication/config path has no per-human organization roles.
- Project roles: Read-only is a server-wide setting, not a named workspace role assigned to a member.
- Sso: Only the instance token/cookie is accepted; no SAML/OIDC or enterprise SSO authentication path exists in this revision.
- Source inspection covers the scoped release only. Unanswered atomic questions remain unknown; no live scenario was run.
-
Agent Deck internal/web/auth.go
Finding: The web server accepts one instance-wide bearer token or matching cookie. Server config has global ReadOnly/WebMutations controls, with no human membership, per-user workspace roles or enterprise SSO provider. Each authenticated terminal WebSocket attaches to the same tmux session and forwards input unless the server is read-only.
SSO: No
SSO
Typed value: No.
Access uses a shared instance bearer token and global read-only mode. The scoped application has no named human workspace roles or enterprise SSO.
- Organization roles: The complete web authentication/config path has no per-human organization roles.
- Project roles: Read-only is a server-wide setting, not a named workspace role assigned to a member.
- Sso: Only the instance token/cookie is accepted; no SAML/OIDC or enterprise SSO authentication path exists in this revision.
- Source inspection covers the scoped release only. Unanswered atomic questions remain unknown; no live scenario was run.
-
Agent Deck internal/web/auth.go
Finding: The web server accepts one instance-wide bearer token or matching cookie. Server config has global ReadOnly/WebMutations controls, with no human membership, per-user workspace roles or enterprise SSO provider. Each authenticated terminal WebSocket attaches to the same tmux session and forwards input unless the server is read-only.
Pull request workflow: No
Pull request workflow
Typed value: No.
Product manages worktrees and branches. The scoped application has no built-in PR creation or management command; a coding agent may run its own external commands.
- Pull requests: No PR creation/management route or command in the inspected client, CLI and Git implementation.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck internal/web/server.go
Finding: The scoped TUI/web routes expose live terminals, session lists, fleet states, files, costs, archive and settings. Git helpers manage branches/worktrees and WIP patches; no ADE-owned diff review interface or pull-request creation/management route is implemented. Agent slash-command passthrough is separate.
Review interface: No
Review interface
Typed value: No.
Transcripts, forks and archives retain session work. The scoped product has no built-in code diff review view.
- Diffs: Inspected TUI and browser views display terminal output and session context, with no product-owned code-diff renderer.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck internal/web/server.go
Finding: The scoped TUI/web routes expose live terminals, session lists, fleet states, files, costs, archive and settings. Git helpers manage branches/worktrees and WIP patches; no ADE-owned diff review interface or pull-request creation/management route is implemented. Agent slash-command passthrough is separate.
Product license: Apache-2.0: No. MIT cross-checked
Product license: Apache-2.0
Typed value: No. MIT.
Agent Deck repository is MIT licensed.
- Client license: MIT.
- Source visibility: Public repository.
- Source revision: bedbb14612fe2445c13af1a0988b42280829dfd6
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: MIT license at pinned application revision.
Product license: MIT: Yes. MIT cross-checked
Product license: MIT
Typed value: Yes. MIT.
Agent Deck repository is MIT licensed.
- Client license: MIT.
- Source visibility: Public repository.
- Source revision: bedbb14612fe2445c13af1a0988b42280829dfd6
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: MIT license at pinned application revision.
Product license: GPL-3.0: No. MIT cross-checked
Product license: GPL-3.0
Typed value: No. MIT.
Agent Deck repository is MIT licensed.
- Client license: MIT.
- Source visibility: Public repository.
- Source revision: bedbb14612fe2445c13af1a0988b42280829dfd6
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: MIT license at pinned application revision.
Product license: GPL-3.0-or-later: No. MIT cross-checked
Product license: GPL-3.0-or-later
Typed value: No. MIT.
Agent Deck repository is MIT licensed.
- Client license: MIT.
- Source visibility: Public repository.
- Source revision: bedbb14612fe2445c13af1a0988b42280829dfd6
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: MIT license at pinned application revision.
Product license: AGPL-3.0: No. MIT cross-checked
Product license: AGPL-3.0
Typed value: No. MIT.
Agent Deck repository is MIT licensed.
- Client license: MIT.
- Source visibility: Public repository.
- Source revision: bedbb14612fe2445c13af1a0988b42280829dfd6
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: MIT license at pinned application revision.
Product license: AGPL-3.0-or-later: No. MIT cross-checked
Product license: AGPL-3.0-or-later
Typed value: No. MIT.
Agent Deck repository is MIT licensed.
- Client license: MIT.
- Source visibility: Public repository.
- Source revision: bedbb14612fe2445c13af1a0988b42280829dfd6
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: MIT license at pinned application revision.
Product license: Elastic License 2.0: No. MIT cross-checked
Product license: Elastic License 2.0
Typed value: No. MIT.
Agent Deck repository is MIT licensed.
- Client license: MIT.
- Source visibility: Public repository.
- Source revision: bedbb14612fe2445c13af1a0988b42280829dfd6
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: MIT license at pinned application revision.
Product license: Custom terms: No. MIT cross-checked
Product license: Custom terms
Typed value: No. MIT.
Agent Deck repository is MIT licensed.
- Client license: MIT.
- Source visibility: Public repository.
- Source revision: bedbb14612fe2445c13af1a0988b42280829dfd6
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: MIT license at pinned application revision.
Product license: Proprietary: No. MIT cross-checked
Product license: Proprietary
Typed value: No. MIT.
Agent Deck repository is MIT licensed.
- Client license: MIT.
- Source visibility: Public repository.
- Source revision: bedbb14612fe2445c13af1a0988b42280829dfd6
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: MIT license at pinned application revision.
Self-hosting: Yes cross-checked
Self-hosting
Typed value: Yes.
Core TUI, web server and conductor run on user-controlled machines.
- Self hosting: Local or SSH server installation.
- Vendor dependencies: Agents and optional chat bridges use separate vendor accounts.
- Formats: TOML config, local session registry and Markdown conductor files.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Telemetry opt-out: Yes cross-checked
Telemetry opt-out
Typed value: Yes.
Usage telemetry is opt-in and has CLI preview/disable controls.
- Telemetry disclosure: Anonymous usage sent to PostHog EU after opt-in.
- Telemetry defaults: Off until one-time consent.
- Telemetry controls: agent-deck telemetry off or DO_NOT_TRACK=1; preview shows payload.
- Provider separation: Website page views and coding-agent requests have separate policies.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Free access: Yes cross-checked
Free access
Typed value: Yes.
MIT source/direct binary installs support local evaluation with own tool accounts. Release v1.16.38 publishes free macOS and Linux binaries on GitHub.
- Ongoing free: Open-source binary/source installation without a paid Agent Deck account.
- Self host cost: Agent usage and own host costs separate.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Starting price: Free: Yes. Free cross-checked
Starting price: Free
Typed value: Yes. Free.
MIT source/direct binary installs support local evaluation with own tool accounts. Release v1.16.38 publishes free macOS and Linux binaries on GitHub.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck v1.16.38 release assets
Finding: The latest release publishes free downloadable macOS and Linux binaries of the CLI/TUI on GitHub.
Starting price: Up to $20/month: No. Free cross-checked
Starting price: Up to $20/month
Typed value: No. Free.
MIT source/direct binary installs support local evaluation with own tool accounts. Release v1.16.38 publishes free macOS and Linux binaries on GitHub.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck v1.16.38 release assets
Finding: The latest release publishes free downloadable macOS and Linux binaries of the CLI/TUI on GitHub.
Starting price: $21–40/month: No. Free cross-checked
Starting price: $21–40/month
Typed value: No. Free.
MIT source/direct binary installs support local evaluation with own tool accounts. Release v1.16.38 publishes free macOS and Linux binaries on GitHub.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck v1.16.38 release assets
Finding: The latest release publishes free downloadable macOS and Linux binaries of the CLI/TUI on GitHub.
Starting price: $41–100/month: No. Free cross-checked
Starting price: $41–100/month
Typed value: No. Free.
MIT source/direct binary installs support local evaluation with own tool accounts. Release v1.16.38 publishes free macOS and Linux binaries on GitHub.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck v1.16.38 release assets
Finding: The latest release publishes free downloadable macOS and Linux binaries of the CLI/TUI on GitHub.
Starting price: Over $100/month: No. Free cross-checked
Starting price: Over $100/month
Typed value: No. Free.
MIT source/direct binary installs support local evaluation with own tool accounts. Release v1.16.38 publishes free macOS and Linux binaries on GitHub.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck v1.16.38 release assets
Finding: The latest release publishes free downloadable macOS and Linux binaries of the CLI/TUI on GitHub.
Starting price: Contact for pricing: No. Free cross-checked
Starting price: Contact for pricing
Typed value: No. Free.
MIT source/direct binary installs support local evaluation with own tool accounts. Release v1.16.38 publishes free macOS and Linux binaries on GitHub.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Agent Deck v1.16.38 release assets
Finding: The latest release publishes free downloadable macOS and Linux binaries of the CLI/TUI on GitHub.
Pricing model: Free, no paid plan: Yes
Pricing model: Free, no paid plan
Typed value: Yes.
Agent Deck is a completely free MIT application with no paid ADE plan in the checked first-party offering. Underlying coding-agent/model and user-owned compute costs are separate.
- First-party pricing pages, terms and licences read 2026-11-10; no purchase, billing or authenticated account test.
-
Finding: The first-party site source declares a 0 USD offer and says Agent Deck is completely free and open source under MIT; install/download/build are the offered distribution routes. No paid ADE tier appears in the offering.
Pricing model: Free plan: No
Pricing model: Free plan
Typed value: No.
Agent Deck is a completely free MIT application with no paid ADE plan in the checked first-party offering. Underlying coding-agent/model and user-owned compute costs are separate.
- First-party pricing pages, terms and licences read 2026-11-10; no purchase, billing or authenticated account test.
-
Finding: The first-party site source declares a 0 USD offer and says Agent Deck is completely free and open source under MIT; install/download/build are the offered distribution routes. No paid ADE tier appears in the offering.
Pricing model: Subscription: No
Pricing model: Subscription
Typed value: No.
Agent Deck is a completely free MIT application with no paid ADE plan in the checked first-party offering. Underlying coding-agent/model and user-owned compute costs are separate.
- First-party pricing pages, terms and licences read 2026-11-10; no purchase, billing or authenticated account test.
-
Finding: The first-party site source declares a 0 USD offer and says Agent Deck is completely free and open source under MIT; install/download/build are the offered distribution routes. No paid ADE tier appears in the offering.
Pricing model: Per seat: No
Pricing model: Per seat
Typed value: No.
Agent Deck is a completely free MIT application with no paid ADE plan in the checked first-party offering. Underlying coding-agent/model and user-owned compute costs are separate.
- First-party pricing pages, terms and licences read 2026-11-10; no purchase, billing or authenticated account test.
-
Finding: The first-party site source declares a 0 USD offer and says Agent Deck is completely free and open source under MIT; install/download/build are the offered distribution routes. No paid ADE tier appears in the offering.
Pricing model: Usage-based: No
Pricing model: Usage-based
Typed value: No.
Agent Deck is a completely free MIT application with no paid ADE plan in the checked first-party offering. Underlying coding-agent/model and user-owned compute costs are separate.
- First-party pricing pages, terms and licences read 2026-11-10; no purchase, billing or authenticated account test.
-
Finding: The first-party site source declares a 0 USD offer and says Agent Deck is completely free and open source under MIT; install/download/build are the offered distribution routes. No paid ADE tier appears in the offering.
Pricing model: Per machine or compute: No
Pricing model: Per machine or compute
Typed value: No.
Agent Deck is a completely free MIT application with no paid ADE plan in the checked first-party offering. Underlying coding-agent/model and user-owned compute costs are separate.
- First-party pricing pages, terms and licences read 2026-11-10; no purchase, billing or authenticated account test.
-
Finding: The first-party site source declares a 0 USD offer and says Agent Deck is completely free and open source under MIT; install/download/build are the offered distribution routes. No paid ADE tier appears in the offering.
Pricing model: Bring your own key: Yes
Pricing model: Bring your own key
Typed value: Yes.
Agent Deck is a completely free MIT application with no paid ADE plan in the checked first-party offering. Underlying coding-agent/model and user-owned compute costs are separate.
- Model cost: Model usage runs through the user's own Claude login/API key or other provider account, billed by that provider; no amount is set by Agent Deck.
- First-party pricing pages, terms and licences read 2026-11-10; no purchase, billing or authenticated account test.
-
Finding: Agent Deck runs the unmodified agent CLIs with the user's own login or API key, so model usage is paid to the provider; the README describes a dashboard that tracks those provider costs but states no Agent Deck price.
Pricing model: One-time purchase: No
Pricing model: One-time purchase
Typed value: No.
Agent Deck is a completely free MIT application with no paid ADE plan in the checked first-party offering. Underlying coding-agent/model and user-owned compute costs are separate.
- First-party pricing pages, terms and licences read 2026-11-10; no purchase, billing or authenticated account test.
-
Finding: The first-party site source declares a 0 USD offer and says Agent Deck is completely free and open source under MIT; install/download/build are the offered distribution routes. No paid ADE tier appears in the offering.
Keeps going with laptop closed: Yes
Keeps going with laptop closed
Typed value: Yes.
Sessions run locally, in Docker or on user-owned SSH/cloud hosts. The scoped execution implementation has no vendor-managed compute service.
- Local: Local tmux sessions.
- Byo remote: remote add/attach/sessions/drain over SSH.
- Self hosted: Own machine runs core app and web UI.
- Other device continuation: Remote host sessions browsed/attached from local deck.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
Open-source license: Yes. MIT cross-checked
Open-source license
Typed value: Yes. MIT.
Agent Deck repository is MIT licensed.
- Client license: MIT.
- Source visibility: Public repository.
- Source revision: bedbb14612fe2445c13af1a0988b42280829dfd6
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: MIT license at pinned application revision.
Any desktop app: Yes
Any desktop app
Typed value: Yes.
CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
- Desktop os: macOS and Linux native terminal binaries; Windows requires WSL, with no native Windows binary.
- Mobile apps: No native iOS or Android application in the scoped release build and client tree; phone access uses the browser.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Release builds target darwin and linux on amd64/arm64. The pinned repository release/build tree contains a Go CLI/TUI and embedded browser UI, with no Windows-native, iOS or Android application target. Windows is documented through WSL in README Installation.
Any phone client: Yes
Any phone client
Typed value: Yes.
CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
- Browser: agent-deck web, with optional read-only mode.
- Mobile browser: Home-screen browser app retains authentication cookie; mobile conductor control also uses chat-service bridges.
- Cli: agent-deck terminal UI and subcommands.
- Task actions: Web terminal can interact with sessions; read-only flag restricts input.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
All public findings
Findings may summarize a source. They are not presented as verbatim statements.
agent-deck:ACC-01: CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
agent-deck:ACC-01
CLI/TUI ships for macOS and Linux; Windows access uses WSL. The released clients are terminal and browser interfaces, with phone browser access and no native phone application in this revision.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
-
Finding: Release builds target darwin and linux on amd64/arm64. The pinned repository release/build tree contains a Go CLI/TUI and embedded browser UI, with no Windows-native, iOS or Android application target. Windows is documented through WSL in README Installation.
agent-deck:ACC-02: Many terminal coding agents are supported with different integration levels. At v1.16.38 an aider session command also maps to a built-in Aider tool identity, without status detection. cross-checked
agent-deck:ACC-02
Many terminal coding agents are supported with different integration levels. At v1.16.38 an aider session command also maps to a built-in Aider tool identity, without status detection.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
-
Agent Deck session tool mapping
Finding: Creating a session with the command aider resolves to the built-in Aider tool identity and runs the aider command in the session.
-
Agent Deck built-in tool registry
Finding: Aider is a registered built-in tool name with its own icon and colour, but carries no detection patterns, so running aider sessions are not status-detected as Aider.
-
Finding: The in-repo website source lists a 0 USD offer, says Agent Deck is completely free and open source under MIT, and says it supports Claude Code, Aider, Gemini CLI, Codex and any terminal program. The page metadata names an old version (0.15.0).
agent-deck:ACC-03: Agent authentication is reused for local/container sessions, with a macOS Claude exception. cross-checked
agent-deck:ACC-03
Agent authentication is reused for local/container sessions, with a macOS Claude exception.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
-
Agent Deck README: Vendor terms and logins
Finding: Agent Deck runs the unmodified claude CLI with the user's own login, recommends an API key or Team/Enterprise plan for heavy unattended use, and supports passing a Claude setup-token via an env_file.
agent-deck:ACC-04: Install binary and add a directory with an installed tool; remote setup can install deck on SSH host. cross-checked
agent-deck:ACC-04
Install binary and add a directory with an installed tool; remote setup can install deck on SSH host.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:ACC-05: CLI exposes session creation, send, stop, fork, parent relationships, MCP, skills, costs and remotes. cross-checked
agent-deck:ACC-05
CLI exposes session creation, send, stop, fork, parent relationships, MCP, skills, costs and remotes.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:ACC-06: Product manages MCP and skills per supported coding tool. cross-checked
agent-deck:ACC-06
Product manages MCP and skills per supported coding tool.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:PLAN-01: Sessions carry semantic goal/state/ticket annotations but external ticket references do not establish native issues. cross-checked
agent-deck:PLAN-01
Sessions carry semantic goal/state/ticket annotations but external ticket references do not establish native issues.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:PLAN-03: Session parent linkage represents delegation, with a documented remote-parent boundary. cross-checked
agent-deck:PLAN-03
Session parent linkage represents delegation, with a documented remote-parent boundary.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:PLAN-04: Conductor identity, shared policy and learnings persist in files. cross-checked
agent-deck:PLAN-04
Conductor identity, shared policy and learnings persist in files.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:PLAN-06: Fleet board cards and work states are session annotations. The scoped persistence and CLI model has no independent native issue/task entity.
agent-deck:PLAN-06
Fleet board cards and work states are session annotations. The scoped persistence and CLI model has no independent native issue/task entity.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
-
Agent Deck internal/statedb/statedb.go
Finding: Persistent state models agent instances, groups, session claims, history, watchers and events. Fleet work states annotate instances; no separate product-owned issue/task entity exists. The run-task command wraps a session worker and records its completion rather than creating an independent issue.
agent-deck:COORD-01: Independent tmux-backed sessions run concurrently and show process/work state separately. cross-checked
agent-deck:COORD-01
Independent tmux-backed sessions run concurrently and show process/work state separately.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:COORD-02: Conductors dispatch and monitor sessions; native issue claims were not established. cross-checked
agent-deck:COORD-02
Conductors dispatch and monitor sessions; native issue claims were not established.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:COORD-03: Persistent conductor sessions orchestrate workers and receive status nudges. cross-checked
agent-deck:COORD-03
Persistent conductor sessions orchestrate workers and receive status nudges.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:COORD-04: Heartbeat/watchers and optional chat-service bridges trigger session work. cross-checked
agent-deck:COORD-04
Heartbeat/watchers and optional chat-service bridges trigger session work.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:COORD-05: Conductor auto-responds or escalates; users can send and stop sessions. cross-checked
agent-deck:COORD-05
Conductor auto-responds or escalates; users can send and stop sessions.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:COORD-06: TUI and web Fleet/Command Center expose grouped session state and attention. cross-checked
agent-deck:COORD-06
TUI and web Fleet/Command Center expose grouped session state and attention.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:COORD-07: Users can work directly in sessions or enable conductors and configure tool/model choices. cross-checked
agent-deck:COORD-07
Users can work directly in sessions or enable conductors and configure tool/model choices.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:RUN-01: Sessions run locally, in Docker or on user-owned SSH/cloud hosts. The scoped execution implementation has no vendor-managed compute service.
agent-deck:RUN-01
Sessions run locally, in Docker or on user-owned SSH/cloud hosts. The scoped execution implementation has no vendor-managed compute service.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
-
Agent Deck internal/session/instance_sandbox_ssh.go
Finding: Session command preparation runs on the local host, user-configured SSH host or Docker container. The scoped wrapper and container launch code provides no vendor-operated execution service or product VM/microVM isolation backend. Cloud provider connectors use the operator account and remain user-controlled infrastructure.
agent-deck:RUN-02: Git worktrees and Docker containers separate execution scope. The scoped session launcher has no VM or microVM isolation backend.
agent-deck:RUN-02
Git worktrees and Docker containers separate execution scope. The scoped session launcher has no VM or microVM isolation backend.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
-
Agent Deck internal/session/instance_sandbox_ssh.go
Finding: Session command preparation runs on the local host, user-configured SSH host or Docker container. The scoped wrapper and container launch code provides no vendor-operated execution service or product VM/microVM isolation backend. Cloud provider connectors use the operator account and remain user-controlled infrastructure.
agent-deck:RUN-03: Session resume/fork support is tool-specific; remote failures are visible and retryable. cross-checked
agent-deck:RUN-03
Session resume/fork support is tool-specific; remote failures are visible and retryable.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:RUN-04: Worktree hooks initialize task environments; Docker setup may require separate authentication. cross-checked
agent-deck:RUN-04
Worktree hooks initialize task environments; Docker setup may require separate authentication.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:RUN-05: Cost budgets and run token/turn/time defaults are configurable, with explicit untested enforcement limits. cross-checked
agent-deck:RUN-05
Cost budgets and run token/turn/time defaults are configurable, with explicit untested enforcement limits.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:RUN-06: Status lights and semantic annotations are separate; roadmap lists detection gaps. cross-checked
agent-deck:RUN-06
Status lights and semantic annotations are separate; roadmap lists detection gaps.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
-
Finding: Unscheduled roadmap lists status detection gaps, remote notification follow-ups and signed/notarized macOS release work. These are not shipped claims.
agent-deck:RUN-07: Session costs and tokens are collected from provider transcripts/output, with limited provider verification. cross-checked
agent-deck:RUN-07
Session costs and tokens are collected from provider transcripts/output, with limited provider verification.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:TEAM-02: Access uses a shared instance bearer token and global read-only mode. The scoped application has no named human workspace roles or enterprise SSO.
agent-deck:TEAM-02
Access uses a shared instance bearer token and global read-only mode. The scoped application has no named human workspace roles or enterprise SSO.
- Source inspection covers the scoped release only. Unanswered atomic questions remain unknown; no live scenario was run.
-
Agent Deck internal/web/auth.go
Finding: The web server accepts one instance-wide bearer token or matching cookie. Server config has global ReadOnly/WebMutations controls, with no human membership, per-user workspace roles or enterprise SSO provider. Each authenticated terminal WebSocket attaches to the same tmux session and forwards input unless the server is read-only.
agent-deck:TEAM-03: Web mode has token and read-only controls; shared token is not organization RBAC. cross-checked
agent-deck:TEAM-03
Web mode has token and read-only controls; shared token is not organization RBAC.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:TEAM-04: Conductor task-log and semantic decisions persist, without an audit integrity guarantee. cross-checked
agent-deck:TEAM-04
Conductor task-log and semantic decisions persist, without an audit integrity guarantee.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:SHIP-01: Transcripts, forks and archives retain session work. The scoped product has no built-in code diff review view.
agent-deck:SHIP-01
Transcripts, forks and archives retain session work. The scoped product has no built-in code diff review view.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
-
Agent Deck internal/web/server.go
Finding: The scoped TUI/web routes expose live terminals, session lists, fleet states, files, costs, archive and settings. Git helpers manage branches/worktrees and WIP patches; no ADE-owned diff review interface or pull-request creation/management route is implemented. Agent slash-command passthrough is separate.
agent-deck:SHIP-03: Product manages worktrees and branches. The scoped application has no built-in PR creation or management command; a coding agent may run its own external commands.
agent-deck:SHIP-03
Product manages worktrees and branches. The scoped application has no built-in PR creation or management command; a coding agent may run its own external commands.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
-
Agent Deck internal/web/server.go
Finding: The scoped TUI/web routes expose live terminals, session lists, fleet states, files, costs, archive and settings. Git helpers manage branches/worktrees and WIP patches; no ADE-owned diff review interface or pull-request creation/management route is implemented. Agent slash-command passthrough is separate.
agent-deck:OWN-01: Agent Deck repository is MIT licensed. cross-checked
agent-deck:OWN-01
Agent Deck repository is MIT licensed.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: MIT license at pinned application revision.
agent-deck:OWN-02: Core TUI, web server and conductor run on user-controlled machines. cross-checked
agent-deck:OWN-02
Core TUI, web server and conductor run on user-controlled machines.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:OWN-03: Usage telemetry is opt-in and has CLI preview/disable controls. cross-checked
agent-deck:OWN-03
Usage telemetry is opt-in and has CLI preview/disable controls.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
agent-deck:OWN-04: Agent Deck is a completely free MIT application with no paid ADE plan in the checked first-party offering. Underlying coding-agent/model and user-owned compute costs are separate.
agent-deck:OWN-04
Agent Deck is a completely free MIT application with no paid ADE plan in the checked first-party offering. Underlying coding-agent/model and user-owned compute costs are separate.
- First-party pricing pages, terms and licences read 2026-11-10; no purchase, billing or authenticated account test.
-
Finding: Agent Deck runs the unmodified agent CLIs with the user's own login or API key, so model usage is paid to the provider; the README describes a dashboard that tracks those provider costs but states no Agent Deck price.
-
Finding: The first-party site source declares a 0 USD offer and says Agent Deck is completely free and open source under MIT; install/download/build are the offered distribution routes. No paid ADE tier appears in the offering.
agent-deck:OWN-06: Public releases and community support are available. cross-checked
agent-deck:OWN-06
Public releases and community support are available.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
-
Finding: Release published 7 November 2026.
agent-deck:OWN-08: MIT source/direct binary installs support local evaluation with own tool accounts. Release v1.16.38 publishes free macOS and Linux binaries on GitHub. cross-checked
agent-deck:OWN-08
MIT source/direct binary installs support local evaluation with own tool accounts. Release v1.16.38 publishes free macOS and Linux binaries on GitHub.
- Documentation establishes only the named facts. Unknown atomic questions and unrun scenarios remain open.
-
Finding: Terminal command center for independent coding sessions, with CLI, web UI, conductor orchestration, Docker/worktrees, remote SSH decks, usage metrics and opt-in telemetry.
-
Agent Deck v1.16.38 release assets
Finding: The latest release publishes free downloadable macOS and Linux binaries of the CLI/TUI on GitHub.