agentsclimarketplace

Idea new

Skill GrinRus/ai_driven_dev/skills/idea-new

Bootstraps a new feature ticket by setting active context, deriving `slug_hint`, and preparing PRD questions. Use when the idea stage starts for a new ticket. Do not use when the request is to refresh research artifacts (`researcher`) or draft implementation plans (`plan-new`).From its SKILL.md

Install
npx -y skills add GrinRus/ai_driven_dev --skill idea-new

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 5 stars5 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

4.5 KB, as published. Nobody here has run it

Follow feature-dev-aidd:aidd-core.

Steps

  1. Inputs: resolve <ticket>, parse idea_note from user text, and verify PRD/context artifacts are readable for idea stage.
  2. Slug synthesis: generate a short LLM summary from idea_note, normalize it into kebab-case token ([a-z0-9-], 2-6 words), and use it as internal slug_hint; if idea_note is empty, fallback to <ticket> token. Do not ask user for slug_hint.
  3. Preflight: set active stage idea and active feature with python3 ${CLAUDE_PLUGIN_ROOT}/skills/aidd-flow-state/runtime/set_active_stage.py and python3 ${CLAUDE_PLUGIN_ROOT}/skills/aidd-flow-state/runtime/set_active_feature.py <ticket> --slug-note "<generated-slug>".
  4. Orchestration: build/update the rolling context pack aidd/reports/context/<ticket>.pack.md and run python3 ${CLAUDE_PLUGIN_ROOT}/skills/idea-new/runtime/analyst_check.py --ticket <ticket>.
  5. Run subagent feature-dev-aidd:analyst. First action: read the rolling context pack. For the persisted PRD artifact, override the global interactive question wording and copy the analyst-dialog labels from templates/prd.template.md verbatim.
  6. Postflight: reject comment-only placeholders as invalid analyst output. Read-time compatibility still accepts transitional Question / Why / Options / Default, but the canonical persisted PRD write format remains the template-native Russian wording.
  7. Postflight: if answers already exist, rerun python3 ${CLAUDE_PLUGIN_ROOT}/skills/idea-new/runtime/analyst_check.py --ticket <ticket>; the validator syncs derived index/status state from the current PRD before returning.
  8. Ready path: return /feature-dev-aidd:researcher <ticket> only when analyst_check.py confirms idea-stage readiness.
  9. Pending path: if required questions remain or PRD fields are incomplete, return PENDING with the full current remaining-question set in the top-level response and keep the next action on idea-new.

Command contracts

python3 ${CLAUDE_PLUGIN_ROOT}/skills/aidd-flow-state/runtime/set_active_feature.py

  • When to run: once ticket and idea note are parsed, before analyst execution.
  • Inputs: <ticket> and generated --slug-note <slug_hint> derived from idea_note.
  • Outputs: updated aidd/docs/.active.json (ticket + internal slug_hint) and PRD scaffold.
  • Failure mode: non-zero exit when workspace paths/permissions are invalid.
  • Next action: fix workspace issue, keep the same generated slug policy, rerun the command.

python3 ${CLAUDE_PLUGIN_ROOT}/skills/idea-new/runtime/analyst_check.py

  • When to run: before/after analyst execution to validate PRD readiness and Q/A synchronization.
  • Inputs: --ticket <ticket> and active workspace artifacts.
  • Outputs: deterministic readiness status for idea-stage gate decisions.
  • Failure mode: non-zero exit when required PRD fields or answer alignment are missing.
  • Next action: update PRD/QA artifacts, then rerun the same validator.

Notes

  • Interactive chat questions still follow the shared aidd-policy format.
  • Persisted PRD artifact language is stage-local here: the analyst dialog must use the exact labels from templates/prd.template.md, and placeholder comments do not count as a valid question block.
  • Non-interactive callers must rely on the latest top-level stage-return as the source of truth for retry payloads; do not rely only on AskUserQuestion panel state or stale example payloads.
  • Planning stage: AIDD:ACTIONS_LOG: n/a.

Additional resources

  • PRD template source: templates/prd.template.md (when: defining or validating required PRD structure; why: keep analyst output aligned with canonical sections).

Keep looking

Skills are one crate of 627,801. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.