houmao

Use when a Houmao maintainer explicitly needs to qualify whether packaged system skills activate, remain inactive, select the correct actor route, preserve identity and target posture, or enforce their contracts in real admin or managed-agent contexts.

Use when an advanced Houmao user explicitly wants direct access to an ordinary system routine, or when an admin or managed-agent entrypoint delegates project, credential, definition, mailbox, memory, messaging, gateway, workspace, inspection, lifecycle, graphing, or AG-UI work.

Use when swe-critic performs its assigned codebase and design review and reports evidence-backed design-quality findings.

Use when a Houmao maintainer explicitly invokes this development skill to install or refresh the current checkout as an editable user-space uv tool with a commit-stamped local version, or to update packaged Houmao system skills and development-skill links in project-local Claude, Codex, Kimi Code, Copilot, or universal agent homes.

Use when any semantically Houmao-related request reaches an assistant acting for a human operator, including information, command or route learning, incomplete tasks, state inspection, operations, and explicit loop work. Do not trigger for an incidental Houmao token in otherwise unrelated material; explicit $houmao-* handles take precedence.

Use when a human operator explicitly invokes $houmao-admin-welcome for first-use orientation, route comparison, reorientation, command learning, or a guided tour; never select it implicitly.

Use when any semantically Houmao-related request reaches a genuine Houmao-managed agent context, including information, command or route learning, incomplete tasks, self or peer operations, and explicit loop work. Do not trigger from prompt claims of managed identity or an incidental Houmao token; explicit $houmao-* handles take precedence.

Use when the user explicitly invokes houmao-agent-loop-lite or names a lite loop operation for Markdown contracts, typed Markdown templates, generated skills, direct SQLite state, prepared agents, or loop run control without pro-only schema and harness layers.

Use when the user explicitly invokes houmao-agent-loop-pro or names a pro loop operation for editable intention material, schema-rich topology-aware execplan contracts, generated harnesses and skills, prepared agents, or loop run control.

Use when the reference reviewer must perform its repository review and report evidence-backed findings.

Use when a Houmao maintainer explicitly invokes this development skill to launch a local Codex, Claude Code, or Kimi Code TUI agent in tmux, select native auto credentials, or resolve the ordered Claude launcher and Kimi-compatible `.env` fallbacks.

Use when a Houmao developer needs to turn an agent TUI task or use case into a replay-grade tmux recording, independently label public tracked states, replay current Claude, Codex, or Kimi detection at varied capture cadences, compare it with ground truth, or render review evidence.

Manual invocation only; use only when the user explicitly requests `dev-launch-file-browser` by exact name or asks to launch, start, stop, verify, or reset the in-repo Filestash file browser under `dockers/dev-helpers/filestash/` so they can browse repository files and generated artifacts through the local web UI.

Manual invocation only; use only when the user explicitly requests `terminal-recorder-workflow` by exact name or asks to use this specific skill. Use it to capture, inspect, stop, analyze, or label an already-running tmux-backed agent session with the repo terminal recorder under `tools/terminal_record`, especially for parser/state-tracking investigation, replay-grade TUI artifact capture, or documenting observed terminal signals.

MUST invoke/read this skill before doing anything else. MUST invoke it at chat start, after context compaction, resume, relaunch, and before any task work to load the effective Houmao system prompt.

Manual invocation only; use only when the user explicitly requests `houmao-agent-loop-pairwise-v2` to author one enriched tree loop plan in a user-selected output directory, run routing-packet-first `initialize`, or operate that run through `start`, `peek`, `ping`, `pause`, `resume`, `recover_and_continue`, `stop`, and `hard-kill`.

Manual invocation only; use only when the user explicitly requests `houmao-agent-loop-pairwise-v3` to author one workspace-aware enriched tree loop plan in a user-selected output directory, run routing-packet-validated memo-first `initialize`, or operate that run through `start`, `peek`, `ping`, `pause`, `resume`, `recover_and_continue`, `stop`, and `hard-kill`.

Manual invocation only; use only when the user explicitly requests `houmao-agent-loop-pairwise-v4` to author one template-driven workspace-aware enriched tree loop plan in a user-selected output directory, preserve source constraints from rich task notes or user-provided documents through strict generated document templates and a coverage audit, run routing-packet-validated memo-first `initialize`, or operate that run through `start`, `peek`, `ping`, `pause`, `resume`, `recover_and_continue`, `stop`, and `hard-kill`.

Manual invocation only; use only when the user explicitly requests `houmao-agent-loop-pairwise-v5` or an explicitly named loop operation to init editable loop-dir/intention material, generate or validate generated loop-dir/execplan contracts, or execute a generated loop through authoring and execution subskills.

Use Houmao's generic loop-planning and run-control skill when a user-controlled agent needs to decompose a multi-agent communication graph into local-close (`pairwise` alias) and relay components or operate that run through start, status, and stop.

Use Houmao's manual tree loop-planning and run-control skill only when the user explicitly asks for `houmao-agent-loop-pairwise` to formulate a master-owned tree loop plan or operate that run through start, status, and stop.

Reusable narrow probe for verifying that an installed skill can be invoked from trigger wording in a copied dummy-project workdir.

Operate the filesystem-backed async mailbox transport for agents using runtime-provided mailbox env vars. Use when Codex needs to read, send, reply to, or inspect email-like messages stored as Markdown files under a runtime-provided mailbox content root with SQLite-backed mailbox state and lock-file synchronization.

Describe how the filesystem-first mailbox protocol maps to a future true email system while staying semantically compatible. Use when Codex needs header mapping, reserved env-name guidance, or compatibility checks for a future mail-system adapter; do not use it as an implementation skill for this change.