lidge-jun/codexclaw / plugins/codexclaw/skills/loop
cxc-loop
UnverifiedUse for PABCD completion loops. Triggers: cxc-loop, continue until done, HOTL, repeated PABCD, 루프 돌려, 끝까지 해줘, docs-first.
cxc-loop — Scoped completion
Intent before activation
Bare operative cxc-loop means scoped HOTL; explicit HITL keeps human P/A/B pauses. Loading, explanation and review activate nothing. Explicit limits win. HOTL adds persistence, never permissions; report missing authority.
A leaf follows its packet, owns no goal/FSM and needs a grant to spawn. Dispatched tasks own their goal/FSM within their packet's grant. Before dispatch, read Dispatch surfaces (DISPATCH-SURFACE-01).
Use cxc-dev for safety, proof, complete selected reads, peer contact, async questions and installed-skill discovery. Select owners by task/class/risk/phase; preserve explicit-only and leaf-safe restrictions and resolve missing mandatory references before work.
Read before the action
| Action | Owner |
|---|---|
| PABCD work or planning | cxc-pabcd |
| HOTL entry, resumption or completion diagnosis | Runtime lifecycle |
| Goalplan schema, CLI or amendment | Durable goalplan |
| Multi-cycle roadmap | Implementation units |
| Repeated failure, reviewer FAIL or archetype selection | Loop engineering |
| Score optimization, plateau or mechanism comparison | Optimization + loop engineering above |
| Candidate divergence | Divergence tiers |
| Task dispatch or lane observation | Lane dispatch |
| Authorized subagent dispatch | Delegation |
| Waiting on work or external processes | Waiting |
| Composition, response projection or in-context computation | Native execution |
Execution invariants
- ORCH-MANDATE-01 (STRICT): read persisted state before claiming loop entry/re-entry; binding and edges follow Phase control (SESSION-IDENTITY-01).
- HOTL needs an ACTIVE host goal and in-flight cycle; HITL needs no goal. Missing capability/binding is a preflight failure, never armed continuation.
- Follow cxc-pabcd's one-work-phase/one-cycle invariant; hooks neither choose nor advance phases.
- LOOP-CONTINUE-01 (DEFAULT): after D or context loss, read the bound goalplan and ledger; preserve criteria and re-enter P while in-scope work remains under the active goal.
- LOOP-UNIT-CHAIN-01 (DEFAULT): add genuinely new in-scope units through a P amendment to the same goal.
- LOOP-CONTINUITY-01 (DEFAULT): next P quotes the previous D conclusion and direction, explains any change, and resumes from durable evidence.
- LOOP-GIT-01 (DEFAULT): before checkpoints or publication, apply cxc-dev's git rules; declared stacks also read Stacked PRs.
Docs-first multi-cycle entry
LOOP-DOCS-FIRST-01 (DEFAULT): this entry is STRICT for HOTL. Two or more work-phases start with a goalplan skeleton and docs-only roadmap cycle. D locks the roadmap; implementation starts next cycle. Later P consumes/revalidates one prewritten decade doc. Scope discovered later pays this debt at next P. No production patches, deploys or implementation-complete claims in the roadmap cycle. Single-cycle work skips it; cxc-dev owns C0/C1 exceptions.
LOOP-READS-PABCD-01 (STRICT): before multi-cycle execution read implementation units above for roadmap contents/numbering; skeletons or empty decade docs do not satisfy it.
Completion and recovery
Report DONE, NOOP, BLOCKED, UNSAFE, NEEDS_HUMAN or BUDGET_EXHAUSTED as outcomes, not FSM phases or host goal statuses. DONE requires cxc-dev's fresh proof and the runtime-lifecycle completion gate (GOAL-COMPLETE-GATE-01); never weaken criteria. Exhaustion requires an actual stated bound. Compaction, wait timeout and Stop release prove neither success nor exhaustion; preserve evidence and recover through the applicable owner.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-changelog-updatemorluto/reaPrepare or rewrite REA release changelogs and GitHub release notes from pinned Git history, with verified contributor thanks and Release Please synchronization. | Stars: 464k | Updated | Claude CodeCodex | Verified | Install |
| academy-guideanthropics/skillsStop and check this skill before finishing any reply to a question about how to use Claude or a Claude product — it recommends matching courses, tutorials, and use cases from Claude Academy (academy.claude.com), Anthropic's learning hub. Trigger on: "how do I", "how can I", "getting started with", "what can Claude do", "teach me", "learn to use"; questions about artifacts, projects, skills, plugins, connectors, MCP; requests about rolling Claude out to a team, class, or organization; and any ask for training materials, onboarding content, or learning resources. Use it when the user is learning how to use a feature or product — not when they are mid-task and just want the task done. This skill composes with other skills: after consulting product documentation to answer how a Claude feature works, also check here for a matching course or tutorial — a docs-grounded answer and an Academy recommendation belong together. Only recommend on a strong match; never invent Academy content. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| claude-apianthropics/skillsReference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent/MCP/tool-definition/multi-agent/RAG/LLM-judge/computer-use; generate/summarize/extract/classify/rewrite/converse over NL; debugging refusals/cutoffs/streaming/tool-calls/tokens). SKIP only when another provider is being worked on (overrides all triggers): OpenAI/GPT/Gemini/Llama/Mistral/Cohere/Ollama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST if no provider named — don't Read the file). | Stars: 196k | Updated | Claude CodeCodex | Unverified | Install |
| doc-coauthoringanthropics/skillsGuide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |