idea-capture

# Idea Capture Use this skill when the user wants to save or update an idea/project discussion. ## Storage - Main idea doc: `ideas/<idea-id>.md` - Session summaries: `ideas/summaries/<idea-id>/<timestamp>.md` - Human index: `ideas/INDEX.md` - Machine catalog: `ideas/catalog.json` ## Inputs Provide what you have: - `title` - `summary` - `notes` - `tags` - `mode`: `create` | `update` | `auto` - optional `idea_id` - optional `source` - optional open questions / next steps ## Matching rules - `update`: require an existing match - `auto`: prefer `idea_id`, then normalized title/slug, else create - avoid duplicate idea files when a clear match exists ## Behavior Use `scripts/idea_capture.py` for the write/update work. Example: ```bash python3 skills/idea-capture/scripts/idea_capture.py \ --title "Desktop Pet OpenClaw" \ --summary "Turn OpenClaw into a desktop pet assistant" \ --notes "Need create/update/session-summary support." \ --tags ai,desktop,agent \ --mode auto \ --source qqbot ``` ## Expected result Report: - idea id - created vs updated - changed files - session summary path ## Guardrails - Keep the main idea doc readable. - Preserve update history. - Put chronology in update logs / session summaries, not in long repeated prose.