$npx -y skills add alexei-led/cc-thingz --skill brainstorming-ideasFollow the base skill. This Claude overlay only defines tool use and platform-specific behavior.
| 1 | # Brainstorming Ideas |
| 2 | |
| 3 | Follow the base skill. This Claude overlay only defines tool use and platform-specific behavior. |
| 4 | |
| 5 | ## Claude tool rules |
| 6 | |
| 7 | - Use `AskUserQuestion` for every choice point. Do not write numbered menus and ask the user to type `1`, `2`, or `3`. |
| 8 | - Ask one question per `AskUserQuestion` call. |
| 9 | - Use single-select for one path, multi-select for multiple risks/constraints/goals, and `allowOther` for custom answers. |
| 10 | - Use `TaskCreate` and `TaskUpdate` to track phases when the session has more than two steps. |
| 11 | - Use `Read`, `Grep`, and `Glob` before asking questions that the repo can answer. |
| 12 | - Use `Write` or `Edit` only after the user approves the exact design note, CONTEXT entry, or ADR target. |
| 13 | - Use external research tools only after the user chooses research or asks for it. |
| 14 | |
| 15 | ## Suggested interactive questions |
| 16 | |
| 17 | Initial question: |
| 18 | |
| 19 | - Header: `Idea type` |
| 20 | - Question: `What would you like to brainstorm?` |
| 21 | - Options: New feature, Modification, Integration, Plan grill/debate, Exploration |
| 22 | - Allow Other: yes |
| 23 | |
| 24 | Assumptions check: |
| 25 | |
| 26 | - Header: `Assumptions` |
| 27 | - Question: `Which assumptions are wrong or risky?` |
| 28 | - Options: All correct, Some wrong, Not sure, Defer this |
| 29 | - Allow Other: yes |
| 30 | |
| 31 | Next-step checkpoint: |
| 32 | |
| 33 | - Header: `Next step` |
| 34 | - Question: `How should we proceed?` |
| 35 | - Options: Explore codebase, Research solutions, Explore then research, Skip to approaches |
| 36 | - Allow Other: yes |
| 37 | |
| 38 | Approach choice: |
| 39 | |
| 40 | - Header: `Approach` |
| 41 | - Question: `Which approach fits best?` |
| 42 | - Options: Recommended option, alternative option, minimal/YAGNI option |
| 43 | - Allow Other: yes |
| 44 | |
| 45 | Design validation: |
| 46 | |
| 47 | - Header: `Validate design` |
| 48 | - Question: `Does this section look right?` |
| 49 | - Options: Yes continue, Needs changes, Go back, Stop here |
| 50 | - Allow Other: yes |
| 51 | |
| 52 | Implementation handoff: |
| 53 | |
| 54 | - Header: `Next steps` |
| 55 | - Question: `What should happen next?` |
| 56 | - Options: Create worktree, Create plan, Save design note, Done for now |
| 57 | - Allow Other: yes |
| 58 | |
| 59 | ## Optional exploration |
| 60 | |
| 61 | If the user chooses codebase exploration, run a bounded read-only scan. Prefer direct `Read`/`Grep`/`Glob`; use a subagent only for broad scans. |
| 62 | |
| 63 | Prompt shape for a broad scan: |
| 64 | |
| 65 | ```text |
| 66 | Quick scan only. Find project structure, relevant flows, conventions, integration points, and tests for: <idea>. Return 5 bullets with file paths. Do not edit. |
| 67 | ``` |
| 68 | |
| 69 | If no relevant code appears, say `no prior implementation found`; do not fabricate patterns. |
| 70 | |
| 71 | ## Optional research |
| 72 | |
| 73 | If the user chooses research, use Perplexity or WebFetch with a scoped query. Summarize sourced patterns before proposing approaches. If live retrieval is unavailable, say so and continue from local context only. |
| 74 | |
| 75 | ## Capture rules |
| 76 | |
| 77 | - For `CONTEXT.md` entries, ask for approval of the exact one-sentence definition before editing. |
| 78 | - For ADRs, require all three: hard to reverse, surprising without context, and a real trade-off. |
| 79 | - For design notes, write only concise Problem, Chosen approach, Trade-offs, Open questions, and Testing strategy. |
| 80 | |
| 81 | ## Failure handling |
| 82 | |
| 83 | - Idea conflicts with domain docs: quote the conflicting terms and resolve with the user before designing. |
| 84 | - A constraint blocks every approach: state the blocker, what would unblock it, and ask what to relax. |
| 85 | - Idea is too vague: stay in understanding mode; ask one narrowing question at a time. |
| 86 | - No bounded plan exists for grill/debate: ask for one; do not invent opposing positions. |
| 87 | - If the user asks for task sequencing, state that it is outside this skill's scope. |
| 88 | - If the user asks for generic technology research, route to `researching-web`. |
| 89 | - If the user asks for docs-only work, route to `documenting-code`. |
| 90 | - If the user stops mid-session, offer `BRAINSTORM PAUSED` or a short design note. |
| 91 | |
| 92 | ### Execute this collaborative brainstorming workflow now |