Stats
Actions
Tags
From writing
Creates and refines reader persona YAML files used by the review pipeline. Use when creating a new reader persona for a project brief, or refining an existing persona based on accumulated review feedback. Do NOT use for running the review itself — use `review` instead; do NOT use to assemble the brief — use `obsidian:brief` instead.
How this skill is triggered — by the user, by Claude, or both
Slash command
/writing:personasonnetThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Manages reader personas consumed by the `review` skill. Two operations: **generate** creates a new YAML persona from a description and writes it to `<brief>/personas/`; **train** refines an existing persona by analyzing patterns across accumulated review feedback files under `<output>/review/` to sharpen scoring criteria, must-haves, and deal-breakers.
Manages reader personas consumed by the review skill. Two operations: generate creates a new YAML persona from a description and writes it to <brief>/personas/; train refines an existing persona by analyzing patterns across accumulated review feedback files under <output>/review/ to sharpen scoring criteria, must-haves, and deal-breakers.
Reads the brief model: ${CLAUDE_PLUGIN_ROOT}/references/brief-model.md.
| # | Action | Role | Input |
|---|---|---|---|
| 01 | generate | Create a new persona YAML from a description or profile | <brief> + persona description |
| 02 | train | Refine an existing persona from accumulated review feedback files | <brief> --out <output> + persona-id |
Trigger-to-action mapping:
generatetraingm-practitioner, casual-reader, fan-wot.<brief>/personas/<id>.yaml.train reads only <output>/review/chapter-<NN>-<persona>.md files; never invent patterns.train fires: only when a persona proves unreliable — it caps (≤11/20) repeatedly (≥3 chapters) on chapters the other personas and the craft checklist judge sound (uncorroborated verdict = the persona drifted). A persona capping on a corroborated real defect is doing its job — fix the text/style, never retrain it (that would silence a valid critic). Trigger defined in ${CLAUDE_PLUGIN_ROOT}/references/review-loop.md.<brief>/ is read-only for all other writing skills; persona is the only skill that writes into <brief>/personas/.${CLAUDE_PLUGIN_ROOT}/references/brief-model.md — the brief → output working-dir contract.${CLAUDE_PLUGIN_ROOT}/references/review-loop.md — convergence loop; defines when train is triggered.<brief>/summary.md — brief autosuffisant (for context if needed).<output>/review/chapter-<NN>-<persona>.md — feedback files consumed by train.npx claudepluginhub rebellioussmile/my-claude-marketplace --plugin writingProvides UI/UX resources: 50+ styles, color palettes, font pairings, guidelines, charts for web/mobile across React, Next.js, Vue, Svelte, Tailwind, React Native, Flutter. Aids planning, building, reviewing interfaces.
Fetches up-to-date documentation from Context7 for libraries and frameworks like React, Next.js, Prisma. Use for setup questions, API references, and code examples.