CANBAN

CANBAN // DOCS // AGENTS

24 Built-in Agent CLIs, Herdr PTYs & CLI

Canban runs native coding agent CLIs inside persistent Herdr PTY terminals (herdr-core) rendered with libghostty-vt, preserving every tool's native TUI, MCP servers, skills, permissions, and conversation resume.

Zero-interference native TUI, MCP & skill preservation

Rather than reimplementing provider chat wrappers or injecting fragile Stop hooks, Canban launches your chosen CLI inside a herdr-core (hc) supervised PTY:

  • Native TUIs & Controls: Interact directly with the agent's own terminal UI via libghostty-vt (rendered in native AppKit/CoreText on macOS and GTK4 on Linux, with SGR/pixel mouse reporting, OSC 8 hyperlinks, and 17 dark themes including Monokai).
  • MCPs, Skills & Permissions Intact: Because the real CLI binary executes in the Card worktree with your home configuration, existing MCP servers, custom skills, sandbox flags, and auth logins work unchanged.
  • Restart-Safe PTYs & Exact Resume: Closing Canban, restarting canband, or live-upgrading the app leaves running herdr-core sessions alive. Finished sessions can resume exact provider conversation IDs (--resume, --session, or --conversation) when supported.
  • Global Scheduler Budget: Automatic workflow steps share a configurable concurrency limit (scheduler.max_agents, default 8), while manual shells and script steps run outside the agent budget.
Canban Card #2 running a native @developer CLI session inside a persistent Herdr PTY terminal alongside card activity, Hold auto-merge, and workflow progress.
Live @developer Herdr PTY running cargo test inside Card #2's isolated Git worktree (acme-api-f4c9/2), waiting for explicit canban report completion.

All 24 built-in CLI agents

Canban's catalog (crates/profiles/src/catalog.rs) includes 24 official interactive CLI presets matching Herdr's terminal detector and resume planner, plus custom profiles in config/agents.json and dynamic herdr:<template> profiles:

IDAgentBinaryExact ResumeModel / Effort Args
codexCodexcodexYes (resume)-m, reasoning effort
claudeClaude CodeclaudeYes (--resume)--model, --effort
agyAntigravityagyYes (--conversation)--model, --effort
geminiGemini CLIgeminiYes (--resume)-m
opencodeOpenCodeopencodeYes (--session)-m, --variant
cursorCursor Agentcursor-agentYes (--resume)--model
piPipiYes (--session)--model, --thinking
ompOmpompYes (--resume=)--model
copilotGitHub CopilotcopilotYes (--resume=)--model
droidDroiddroidYes (--resume)-m
ampAmpampFresh launch—
kimiKimikimiYes (--session)--model
kiroKirokiro-cliFresh launch--model
qwenQwen CodeqwenYes (--resume)-m
devinDevindevinYes (--resume)—
clineClineclineFresh launch—
grokGrokgrokYes (--resume)-m
hermesHermeshermesYes (--resume)-m
kiloKilokiloYes (--session)-m
qodercliQoderqodercliYes (--resume)--model
lettaLettalettaYes (--conversation)-m
mastracodeMastra CodemastracodeYes (--thread)--model
makiMakimakiFresh launch—
museMusemuseFresh launch—

Interactive takeover with canban card adopt

When you start exploring a Backlog Card in an interactive terminal or @mention session and decide to move ahead with implementation in that same conversation, the agent does not need to lose its context or spawn a duplicate session:

sh
"$CANBAN_CLI" card update --file "$CANBAN_RUN_DIR/card.md"
"$CANBAN_CLI" card adopt --stage develop
  • card adopt automatically passes inherited HC_TERMINAL_ID and HC_SESSION context to AppServer.
  • AppServer verifies the Core, machine, live PTY, Profile, and Card workspace, transitions the Card to develop (reason: "agent_adopted_stage") without launching a second Agent, and binds the existing live PTY directly to the WorkflowRun step.
  • When the agent later runs canban report --state completed, the workflow proceeds seamlessly to Review and Queue Merge.

Injected canban CLI reference

Canban injects CANBAN_CLI, CANBAN_ENDPOINT, CANBAN_HOME, CANBAN_CARD, CARD_ID, CANBAN_PROJECT, CANBAN_PROFILE_ID, and CANBAN_RUN_DIR into every launched session:

sh
# Inspect or update the current Card
"$CANBAN_CLI" card show --fields number,title,body,stage,revision
"$CANBAN_CLI" card update [--title 'Updated title'] --file "$CANBAN_RUN_DIR/card.md"

# Report execution outcome (completed | backlog | blocked | failed | working)
"$CANBAN_CLI" report --state completed --message 'Summary' --file "$CANBAN_RUN_DIR/report.md"

# Reply to a follow-up question without ending the active task
"$CANBAN_CLI" reply --to <QUESTION_ID> --file "$CANBAN_RUN_DIR/reply.md"

# Stage local diff review comments without publishing directly to GitHub
"$CANBAN_CLI" review draft --file "$CANBAN_RUN_DIR/review_draft.json"

# Call any AppServer JSON-RPC method
"$CANBAN_CLI" call card.list '{"project_id":"<PROJECT_ID>","stage":"develop","limit":50}'

For the full command matrix, Agent-first --fields / request_key contract, and --human examples, see Canban CLI, Agent Optimization & --human.