--- name: skill-curator description: Use when researching books, movies, talks, articles, or notes as source material and turning them into draft source-inspired public skills. Also use when choosing how many routing layers a curated skill needs (flat, single-layer, or two-level routing). metadata: internal: true --- # Skill Curator ## Overview Turn source material into practical agent skills. A source is raw material; the published skill taxonomy is organized by capability pack, not by book, movie, author, character, or title. Each curated skill takes one of three shapes — flat, single-layer (hub-and-spoke), or two-level routing — depending on how much internal branching the content needs. Pick the shape explicitly with `references/depth-rubric.md`; do not default to whichever shape feels familiar. ## Operating Contract - Use web research from multiple source types before drafting public skills. - Keep research and provenance detail in `.agents/state/`. - Write public draft skills only under `skills//`. Capability pack is recorded in `skill.json.tags`, not as a directory layer. - Keep `SKILL.md` runtime-only. Put user-facing provenance in `skill.json`. - Pick the skill's shape using `references/depth-rubric.md`. Then scaffold the file set from the matching `references/shapes/.md`. - Use `skill.json.inspired_by` for concise provenance and add a concise grounding reference only when source-to-heuristic mapping will materially help future agents apply the skill. - Prefer source-inspired behavior over source summaries. - Do not reproduce long copyrighted passages, chapter summaries, scripts, dialogue, lyrics, or distinctive text. Paraphrase concepts into operational methods. - Set every generated public `skill.json.status` to `draft`. ## Activation Handshake If the user gives a concrete source and asks for curation, proceed in Guided Draft mode unless they ask for another mode. If the user only invokes this skill, ask for one source seed: - source title, creator, or URL - target audience if known - preferred output: proposal only or draft files ## Modes - **Grill Me**: ask one open question at a time before research when the desired audience, safety boundaries, or shape strategy is unclear. - **Guided Draft**: default. Research, propose the shape and plan, ask one approval question, then write draft files if approved. - **Autopilot**: research and create draft files using conservative assumptions. Stop only for legal/safety ambiguity, unavailable research, or destructive actions. ## Workflow 1. Load `references/use-case-registry.csv`. 2. Load the relevant detail files for the source type and requested output. 3. Inspect the repo's current public skills to align with existing conventions (`skill.json` schema, eval patterns, path layout). Read `skills/dx-heuristics/` as the canonical example for the two-level shape. 4. Build or update a source dossier in `.agents/state/source-dossiers/`. 5. Extract reusable behaviors, anti-patterns, rubrics, workflows, and examples from the source. 6. For each candidate, decide whether it is a small skill, workflow skill, reference addition, or capability-pack-level pattern. 7. Tag each candidate with its capability pack (a `tags` entry in `skill.json`, not a directory). 8. **Pick the shape (depth)** for each candidate using `references/depth-rubric.md`: `flat`, `single-layer` (hub-and-spoke), or `two-level` (intent × surface routing). Bias toward the shallowest shape that fits — a skill can be promoted later, but rarely flattened. 9. If writing files, scaffold `skills//` from the matching `references/shapes/.md`. Load only that shape's anatomy — do not over-build. 10. For shapes that include a registry, map every reference file through it. No public reference should be orphaned or reachable only by scanning a directory. 11. Keep public grounding concise: source detail belongs in the private dossier, user-facing provenance belongs in `skill.json`, and any public grounding reference should map source families to operational heuristics or caveats. 12. Run `just check` (which executes the repo's static checks across all skills). 13. Hand off to `skill-reviewer` before any draft becomes reviewed or published. ## Pack Rules A capability pack is a categorization like coaching, decision making, communication, attention, reflection, or planning. Packs travel in `skill.json.tags` — they are not directory layers in this repo. Never use a source title as a pack tag. Add a new pack tag when: - the name describes a capability domain, not a source - the domain would still make sense if the source disappeared - the pack can accept future sources from books, movies, notes, talks, or articles Use an existing pack tag when: - a current pack already describes the user need - the candidate extends an existing skill's references, rubric, or templates - a new pack would differ only by wording See `references/pack-placement-rubric.md` for the full candidate-shape rubric and decision recording format. ## Public File Rules Every generated public skill must include: - `SKILL.md` — agent runtime instructions only. - `skill.json` — user-facing catalog metadata. Schema follows existing skills in this repo; see `skills/dx-heuristics/skill.json` for the worked example with `inspired_by` as an object array. The rest of the file set depends on the chosen shape — see the matching file in `references/shapes/`. In short: - **Flat** — typically just `SKILL.md`, optionally one or two supporting files. - **Single-layer (hub-and-spoke)** — adds `references/`, optional `templates/`, optional `evals/`, with a single registry CSV routing the use cases. - **Two-level routing** — adds `references/intent-router.csv`, `references/intents/.csv`, `references/playbooks/.md`, shared `references/core/` rubrics, `templates/.md`, and `evals/` including a static-check script. Do not put author biographies, further-reading sections, source marketing, or long bibliographies in `SKILL.md`. If public grounding is useful, keep it short, paraphrased, and routed through the skill's own registry. ## Reference Map - `references/use-case-registry.csv` — curator's own use-case routing. - `references/research-dossier-playbook.md` — web research and dossier rules. - `references/pack-placement-rubric.md` — capability-pack categorization. - `references/depth-rubric.md` — choose flat / single-layer / two-level. - `references/shapes/flat.md` — anatomy for the simplest shape. - `references/shapes/single-layer.md` — anatomy for hub-and-spoke routing. - `references/shapes/two-level.md` — anatomy for intent × surface routing (matches `skills/dx-heuristics/`). - `references/draft-skill-playbook.md` — cross-shape rules (path, `SKILL.md`/`skill.json` separation, evals, quality bar). - `templates/source-dossier.md` — local research dossier template. - `templates/candidate-plan.md` — curator proposal template.