Convert product ideas into schema-aligned discover-notes.md for downstream specification.
npx skills add https://github.com/bpawlakj/ai-devkit --skill discover-bpawlakj
Please help me install this Agent Skill. Skill: discover Source: https://github.com/bpawlakj/ai-devkit/tree/main/claude/skills/discover Command: npx skills add https://github.com/bpawlakj/ai-devkit --skill discover-bpawlakj
It structures an early product discovery conversation so your inputs become schema-aligned docs/discover-notes.md that /product-spec can reliably transform into a full spec, whether you are starting fresh or changing an existing system.
docs/discover-notes.md and resumes from the next unfinished phase without replaying completed work.docs/product-spec.md, docs/architecture/, and docs/work/) to inherit stable baseline elements and ask only what is changing.docs/reference/ and threads it through phases as check-against context, not silent authority.references/product-spec-schema.md for the required checkpoint format so downstream generation is consistent.Run the discover skill with your idea inline, for example: /discover a lightweight inventory app that suggests reorder items from current stock.
High-intent search queries and answers about installing and using this skill.
Structured product discovery converts your idea into schema-aligned discover notes using a phase-based facilitation workflow. It guides context detection and schema checkpointing to ensure your inputs become reliable downstream specifications.
Brownfield inheritance scanning extracts stable baseline elements from existing artifacts like product specs and architecture docs. Delta gating then asks only about changing components, avoiding redundant questioning during discovery.
Resume handling detects prior sessions in your discover notes file and continues from the next unfinished phase. It avoids replaying completed work, making requirements elicitation fully resumable.
Schema checkpointing enforces a consistent format during discover note generation. This structured output allows downstream product-spec generation to reliably transform your elicited requirements.
Reference materials are optional during product discovery. You can select evidence from a docs directory to thread through phases as check-against context, but the workflow uses only user-provided content as authoritative input.
rambozRun a structured 13-section Q&A to populate project vision and architecture docs.
kryptonum-devConvert product ideas into schema-aligned shape-notes.md for PRD generation.
dxc-dannySchedule a code-review sub-agent to evaluate changes between BASE_SHA and HEAD_SHA.
srobinsonCreates a structured SPEC.md template through interactive one-question-at-a-time elicitation.