Explorar skills

Use PROACTIVELY when the user invokes /implement <slug> or asks "now build it / run the tasks / execute the plan". Runs the `implementer` subagent in a forked context which walks specs/<slug>/tasks.md and writes src/** + tests/** one task at a time, one commit per task.

Use PROACTIVELY when the user invokes /plan <slug> or asks "how should we build this". Runs the `planner` subagent in a forked context which writes specs/<slug>/plan.md from a clarified spec. Refuses to plan if spec.md is still `draft`.

Use PROACTIVELY after /plan, /tasks, or /implement completes — and any time the user invokes /review <slug> [artefact]. Runs both the `verifier` subagent (forked context, Claude-native) and codex (via `codex exec`) on the named artefact. Writes specs/<slug>/reviews/{verifier,codex}-<artefact>-<date>.md. Never edits the implementation.

Use PROACTIVELY when the user invokes /clarify <slug> or when a spec has Open questions and the user wants to resolve them. Reads specs/<slug>/spec.md §6, asks targeted questions, integrates answers, bumps Status to `clarified`.

Use PROACTIVELY when the user wants to author, amend, or revise memory/constitution.md (the supreme project contract). Triggers on /constitution and on phrases like "set the project principles", "redéfinis les principes", "what are our invariants". Never use during a feature flow (specify/plan/...) — pause the flow and explain.

Use PROACTIVELY when the user invokes /specify <slug> or asks to "write the spec", "define the feature", "what should we build". Runs the `specifier` subagent in a forked context which writes specs/<slug>/spec.md. Refuses to write code or plan.

Use PROACTIVELY when the user invokes /tasks <slug> or asks "break this plan into tasks". Runs the `decomposer` subagent in a forked context which writes specs/<slug>/tasks.md. Refuses if plan.md is still `draft`.

[PM Twin] Full productize suite — Intelligence Hub, production features, hero page, help docs, sales toolkit.

[PM Twin] Generate or update product roadmap from PRD phases and feature index.

Optimize JavaScript library size.

Create and manage Claude Code skills following Anthropic best practices. Use when creating new skills, modifying skill-rules.json, understanding trigger patterns, working with hooks, debugging skill activation, or implementing progressive disclosure. Covers skill structure, YAML frontmatter, trigger types (keywords, intent patterns, file paths, content patterns), enforcement levels (block, suggest

Trigger on "package for exec", "consult deliverable pack", "appendix vs main deck", "client-ready consulting doc", "strip placeholders proposal", "readout structure". Assembles exec layer vs detail, removes placeholders, runs consistency pass. Produces client-facing artifact + lint report (inline `consult-consistency-lint` or separate).