archon-cli

Interactive first-run wizard walking the user through 8 configuration prompts and optionally bootstrapping the project. Use on first launch or when reconfiguring archon-cli.

Meta-skill that helps users author new SKILL.md skills with proper structure, progressive disclosure, and bundled resources. Use when user wants to create, write, or build a new skill for archon-cli.

Adjust behavioral rules for a custom agent or globally. Uses LLM-mediated merging with diff validation to prevent hallucinated rule changes.

Create a custom agent from a natural language description. Generates agent definition files (agent.md, context.md, tools.md, behavior.md, memory-keys.json, meta.json) in .claude/agents/custom/{name}/.

Review and apply evolution suggestions for a custom agent. Supports FIX (in-place repair), DERIVED (create variant), and CAPTURED (extract from success).

Rollback behavioral rules for an agent to a previous version. Creates a new version (never deletes history).

Invoke a custom agent with a task description. Assembles the Context Envelope from agent definition files, memory, and LEANN, then spawns a Task tool subagent.

Run scripts/ci-gate.sh, surface findings inline, and suggest /diagnose for any failing gate. Use when you want to check CI gate status or verify all dev-flow gates pass.

Sequential one-command run of /to-prd → /prd-to-spec → /spec-to-tasks, then hands off to the user for manual /archon-code invocation. Use when you want the full PRD-driven development loop in one command.

Refine the tasks/phase<N>/task<M>.md tree from /prd-to-spec into atomic, dev-flow-ready task files with verification checklists. Use after /prd-to-spec to validate and refine the generated task specs.

Disciplined diagnosis loop for hard bugs and performance regressions. Reproduce → minimise → hypothesise → instrument → fix → regression-test. Use when user says "diagnose this" / "debug this", reports a bug, says something is broken/throwing/failing, or describes a performance regression.

Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me".

Grilling session that challenges your plan against the existing domain model, sharpens terminology, and updates documentation (CONTEXT.md, ADRs) inline as decisions crystallise. Use when user wants to stress-test a plan against their project's language and documented decisions.

Test-driven development with red-green-refactor loop. Use when user wants to build features or fix bugs using TDD, mentions "red-green-refactor", wants integration tests, or asks for test-first development.

Tell the agent to zoom out and give broader context or a higher-level perspective. Use when you're unfamiliar with a section of code or need to understand how it fits into the bigger picture.

Show the version history and evolution lineage for a custom agent. Supports rollback to any previous version.

Archive a custom agent (move to archived directory) or restore a previously archived agent.

List all custom agents with metadata (creation date, last used, invocation count). Supports --verbose and --all flags.