claude-plugins

Dispatch and collect the finding-verifier fleet at stage_23_verify_findings (PLN-722). Reads verify_manifest.json (written by stage_22b_verify_prepare), spawns one falsify-oriented verifier Task per to_verify[] entry with mode-specific Task scheduling (GitHub mode dispatches verifiers synchronously; local mode uses parallel background + blocking TaskOutput), skips cache_hits[], and collects outputs without retry — missing outputs degrade to pending_verification[] (and, in GitHub mode, a missing BLOCKING/HIGH verifier raises a coverage-gap signal). Invoke when the walker reaches stage_23_verify_findings. Do NOT use for the reviewer fleet (stage_20 — see the spawn-reviewers skill) or the PLN-725 singletons (stage_11/stage_15 — see the singleton-dispatch skill).

Spawn and collect the reviewer fleet at stage_20_spawn_reviewers. Consumes spawn.json.spec (the authoritative spawn spec from derive-spawn-spec / derive-static-spec), resolves GRAPH_PROJECT, builds per-agent prompts from the per-agent template + role suffixes (Bug Hunter A/B, Unified Auditor, Domain Critics, Impact Analyzer), handles the standard, fast-path, all-cached-BHA, and gated-by-verify cases, and runs the spawn/collection contract and agent-failure recovery. Falls back to the static reviewer table when spawn.json marks arbitrate_status:"fallback". Invoke when stage_20_spawn_reviewers is reached (both MODE=local and MODE=github). Do NOT use for the verifier fleet (stage_23 — see the verify-findings skill) or the PLN-725 singletons (stage_11/stage_15 — see the singleton-dispatch skill).

PLN-725 single-agent dispatch for stage_11_extract_signals and stage_15_coverage_critic. After each prepare stage writes its manifest, this skill reads the status field (cache_hit / skipped / needs_agent) and, only on needs_agent, spawns one synchronous singleton Task that writes the by-convention pln725_*.json output the sibling consolidate stage consumes. Invoke from walker-contract step 6 immediately after stage_11_extract_signals or stage_15_coverage_critic finishes. Do NOT use for the reviewer fleet (stage_20 — see the spawn-reviewers skill) or the verifier fleet (stage_23 — see the verify-findings skill).

Verify and fix BLOCKING/HIGH code review findings from a prior review session via category-dispatch, then run project verification.

Render the operator-facing local-mode code review results at stage_29_present. Covers BLOCKING/HIGH/MEDIUM section templates, Justified Findings (PLN-721), Dismissed Findings (PLN-722), Verifier Stats footer (PLN-773), operator-flag descriptions, override precedence rule for stage_22b, Validation Summary, and final Summary. Invoke when MODE=local AND stage_29_present is reached. Do NOT use for GitHub mode — see prompts/github-review.md. Do NOT use for Gate A hygiene-only early-exit — that path is mode-agnostic and remains in start.md alongside the Gate A definition.

Use when the user wants a code-grounded decision table for current behavior, wants to compare current behavior against a plan or work item, or needs a control-flow artifact for recovery, retry, finalization, validation, state-machine, or review-heavy edge cases.

Use to run the Claude Design to ClosedLoop pipeline against the current web-ui. Stage A inventories a design export zip into schema-validated findings (typed design units - screens, regions like nav bars, standalone components like a chat dialog; UX and behavioral changes; Storybook component reuse mapping; token drift vs the live design system), then creates a platform "Design Review" Feature document the team reviews by editing. Stage B is that in-document human review (delete a section to decline, edit a line to amend, leave to accept). Stage C derives decisions from the edited document and generates DRAFT feature tickets grouped per screen (UI plus optional API) for accepted work only. Triggers on "design inventory", "parse claude design export", "design handoff report", "what changed in this design", "generate tickets from design review".

Canonical reusable orchestration procedures for the ClosedLoop planning/implementation orchestrators — PLAN_VALIDATION_SEQUENCE and AWAITING_USER_SEQUENCE. Single source of truth shared by prompts/prompt.md (full loop), prompts/plan-prompt.md (/code:create-plan) and prompts/execute-prompt.md (/code:execute-implementation), which each load standalone and cannot reference one another at runtime. Triggers: - An orchestrator phase needs full plan validation (structural + semantic) — run PLAN_VALIDATION_SEQUENCE. - An orchestrator reaches a hard stop that requires user action before continuing — run AWAITING_USER_SEQUENCE.

Deterministic plan.json validation via Python script, replacing most plan-validator agent calls. Performs JSON parsing, schema validation, task checkbox regex, required section checks, sync validation, and data extraction. Only semantic consistency checks (storage/query alignment) require the LLM agent. Triggers on: plan validation, checking plan format, extracting plan data. Returns PLAN_VALID with extracted data or PLAN_FORMAT_ISSUES with issues list.

Compresses artifacts for judge evaluation. Reads a single raw artifact, applies tiered summarization within a token budget, and returns compacted content with metadata. Isolation via forked context prevents pollution of agent context

Orchestrate parallel judge agent execution, aggregate CaseScore results, write plan-judges.json, code-judges.json, prd-judges.json, or feature-judges.json, and validate output. Supports evaluating implementation plans (16 judges, 4 batches), code artifacts (11 judges, 3 batches), PRD artifacts (5 judges, 2 batches), or Feature artifacts (3 judges, 1 batch) via --artifact-type parameter.

Skip Phase 7 rebuild when no code changed since Phase 5 build passed. Compares git diff hash against stored hash from last successful build validation. Triggers on: entering Phase 7, checking build status, before final build validation. Returns BUILD_CACHE_HIT to skip or BUILD_CACHE_MISS to re-run build-validator.

Run Codex to review a plan file and return structured feedback with a verdict. Triggers on: debate loop Codex review rounds. Returns VERDICT:APPROVED or VERDICT:NEEDS_CHANGES plus CODEX_SESSION token.

Check if critic reviews are still valid before re-running Phase 2.5 critics. Compares plan.json + critic-gates.json content hash against stored hash from last critic run. Triggers on: entering Phase 2.5, checking critic cache, before launching critics. Returns CRITIC_CACHE_HIT to skip critics or CRITIC_CACHE_MISS to re-run them.

Check if cross-repo coordinator results can be reused, avoiding redundant Sonnet agent launches. Compares peer repo git hashes against stored hashes from last coordinator run. Triggers on: entering Phase 1.4.1, checking cross-repo cache, before discovering peers. Returns CROSS_REPO_CACHE_HIT with cached status or CROSS_REPO_CACHE_MISS to re-run coordinator.

Sync plan.md with plan.json content. MUST be used after ANY edit to plan.json to keep plan.md in sync. Triggers: - After editing plan.json with Edit tool (REQUIRED - always sync after edits) - User asks to "sync plan.md", "update plan.md", "extract the md from a plan" - Converting plan.json to markdown or viewing plan as standalone markdown IMPORTANT: Whenever plan.json is modified, plan.md must be updated to match.

This skill should be used when needing to locate files within the Claude Code plugins cache directory (~/.claude/plugins/cache). Triggers include finding tool scripts, skill files, or any plugin resource when the hardcoded path is unknown or varies by plugin version. Use when slash commands or orchestrators need to dynamically resolve plugin file paths.

Check for a cached plan-evaluation.json result before launching the plan-evaluator agent. This skill should be used in Phase 1.3 (Simple Mode Evaluation) of the orchestrator prompt. Triggers on: entering Phase 1.3, checking simple mode, evaluating plan complexity. Returns EVAL_CACHE_HIT with cached values or EVAL_CACHE_MISS signaling re-evaluation is needed.

Upload a file as a ClosedLoop document (PRD, implementation plan, feature, or template). Reads file content and uploads via MCP without consuming conversation context. Also supports creating new versions of existing documents. Triggers on: "upload artifact", "upload PRD", "upload implementation plan", "upload feature", "create artifact from file", "save as artifact", "push to closedloop", "new artifact version", "test artifact upload", "verify artifact content", "upload to project".

This skill should be used when designing prompts, system prompts, or context windows for Claude. Triggers include writing prompts for API calls, designing agent instructions, structuring complex inputs, optimizing context for accuracy, using examples effectively, or implementing chain-of-thought reasoning. Provides comprehensive guidance from Anthropic's official prompt engineering documentation.

Protocol for iteratively refining sub-agent queries through follow-up questions to ensure sufficient context

Provides ClosedLoop environment paths (CLOSEDLOOP_WORKDIR, CLAUDE_PLUGIN_ROOT) to agents. This skill should be used by any agent that needs to access ClosedLoop run directories, plugin schemas, or other path-dependent resources.

Guide for creating NEW skills from scratch. This skill should be used when users want to create a new skill or scaffold a skill directory. Triggers on "create a skill", "new skill", or "scaffold skill". Do NOT use for updating existing skills - use claude-code-expert instead.

This skill helps PMs draft lightweight PRDs. It should be used when a PM wants to define a new feature, brainstorm requirements, or prep for sprint planning. Triggers include "I have a feature idea", "help me write a PRD", "let's break this into stories", or "what do I need for sprint planning". Optimized for speed — generates minimal viable documentation through conversation, not heavy templates.

Structured format for capturing high-quality learnings during ClosedLoop runs

Conventions for editing plan.json implementation plans including task format, structured arrays, and plan structure rules. Use when creating or modifying plan.json files.

Used when creating or updating implementation plans.

This skill should be used when working with Claude Code infrastructure including agents, skills, slash commands, hooks, plugins, settings.json, or CLAUDE.md files. Triggers on YAML frontmatter, .claude/ directory files, Task tool configuration, or hook debugging. Do NOT use for creating NEW skills from scratch - use claude-creator instead.

This skill should be used when a user asks to explain a complex idea, concept, or system architecture, or when a diagram would be helpful to visualize control flows, system architectures, data flows, state machines, sequence diagrams, or entity relationships.

This skill should be used when writing or parsing TOON (Token-Oriented Object Notation) files. Triggers on .toon file operations, org-patterns.toon generation, or when converting between JSON and TOON formats. Provides syntax rules, quoting conventions, and examples for LLM-optimized data serialization.