菜单
Use when starting work in the IronClaw repo, deciding where a feature/fix/prompt/doc belongs, tracing how a request flows, looking up which crate owns a subsystem, or when repo docs, the knowledge graph, or component names seem stale, missing, or contradictory.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when adding or reviewing tests for Reborn behavior — choosing a test tier, covering a bug fix, testing model/tool-choice behavior, touching tests/integration or tests/fixtures/llm_traces, or when a test needs Postgres, Docker, or a live LLM.
Navigate building a user-facing feature in the Reborn stack (a capability that crosses product_workflow → composition → webui_v2 → runtime/serve → frontend). Use when planning or implementing any new Reborn settings page, endpoint, facade method, or runtime-backed capability — especially before writing code, to avoid rebuilding what already exists and to wire it in one pass instead of layer-by-layer.
Use when asked to "review the open PRs", review a batch or stack of pull requests, or run a recurring PR-review pass on a repo — especially with many PRs, stacked branches, conflicts, or security-sensitive changes. Covers grouping, fan-out to review subagents, verdict synthesis, and posting.
Generate or update the IronClaw architecture overview video using Remotion. Use when asked to update, regenerate, or modify the architecture video, add/remove scenes, or reflect codebase changes in the video.
Use when writing or reviewing a change in crates/ that adds a trait, a crate, a dependency edge, a re-export, or code in ironclaw_reborn_composition — or when deciding whether an abstraction, layer, or crate boundary is justified in the IronClaw Reborn stack.
Use when creating or editing agent guidance in this repo — anything under .claude/ (skills, commands, rules), AGENTS.md or CLAUDE.md files, or the runtime skills/ directory — or when guidance is found citing files, branches, or checks that no longer exist.
| name | ironclaw-reborn-orientation |
| description | Use when starting work in the IronClaw repo, deciding where a feature/fix/prompt/doc belongs, tracing how a request flows, looking up which crate owns a subsystem, or when repo docs, the knowledge graph, or component names seem stale, missing, or contradictory. |
The corrected map, verified against HEAD on 2026-07-02 by a full architecture audit. The repo's own docs drift; when this skill and a doc disagree, re-verify with the greps given here.
crates/ — never src/ (v1 monolith, being strangled; nothing depends on it). Binary: ironclaw-reborn, built from crates/ironclaw_reborn_cli (root docs sometimes say "reborn_cli"). For any endpoint/facade/capability work, use the reborn-feature skill.webui_v2_static) → route+handler (webui_v2) → RebornServicesApi facade (product_workflow) → port impl (reborn_composition) → substrates. Turn execution: SessionThreadService (threads) → TurnCoordinator (turns) → TurnRunScheduler (host_runtime) claims → RebornTurnRunExecutor (reborn) → PlannedDriver → CanonicalAgentLoopExecutor (agent_loop) → host ports (loop_support) → CapabilityHost (capabilities) → dispatcher → runtime lanes. Boot: commands/serve.rs → build_reborn_runtime → build_webui_services.TurnRunScheduler + RebornTurnRunExecutor; expired lease is terminal Failed{lease_expired} (RecoveryRequired is a legacy status).ironclaw_reborn_webui_ingress (WebChat's host side). Do not default it into ironclaw_reborn_composition — that crate is for assembly, and existing Slack/auth host code there is debt, not precedent..md file inside the owning Reborn crate, loaded via include_str! (pattern: crates/ironclaw_loop_support, crates/ironclaw_turns, crates/ironclaw_skill_learning prompt files). crates/ironclaw_engine/prompts/ is the v1 engine's directory — root CLAUDE.md still points there; don't follow it for new Reborn work.Five crates in crates/ serve only the v1 monolith: ironclaw_engine (a complete second agent loop — boundary tests forbid Reborn crates from using it), ironclaw_tui, ironclaw_gateway, ironclaw_oauth, ironclaw_embeddings. Verify any crate's status before building on it: python3 -c ... is overkill — just check who depends on it: grep -rl "<crate_name>" crates/*/Cargo.toml Cargo.toml. If the only consumer is the root package, it's v1-only.
bash scripts/codebase-graph.sh status — once. If FRESH and the codebase-memory MCP tools are available, use graph recipes (CLAUDE.md).crates/AGENTS.md routing table + targeted grep. Note the fallback in your report; do not retry the graph.AGENTS.md/CLAUDE.md first; when absent, fall back to CONTRACT.md/README.md, Cargo.toml, and the primary src/ entrypoint. Treat the crates/AGENTS.md map as a routing aid, not proof that an omitted crate is irrelevant.openwiki/ is generated narrative — read for what/why, never edit (regenerated by CI)..claude/skills/ = developer workflow (this skill). Top-level skills/ = product runtime skills compiled into the shipping binaries. Editing skills/ changes the product. See ironclaw-reborn-skill-maintainer before touching either.
reborn-feature (build a feature) · ironclaw-reborn-architecture-review (boundaries/abstractions) · ironclaw-reborn-testing (tiers/harness) · ironclaw-reborn-skill-maintainer (guidance hygiene).