skills/team-brainstorm/SKILL.md
Optional collaborative brainstorming phase before research — explores user intent, requirements, and design alternatives interactively. Trigger on "/team-brainstorm <idea>" or prefix the full pipeline with brainstorming by running this before /team.
npx skillsauth add bostonaholic/team team-brainstormInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Run a collaborative brainstorming session to explore a feature idea before committing to implementation. This phase is interactive — it asks questions, surfaces alternatives, and produces a refined feature specification that the full TEAM pipeline can act on.
Use this when:
Skip this phase when:
[BRAINSTORM] → RESEARCH → PLAN → TEST → IMPLEMENT → VERIFY → SHIP
Brainstorming is optional. It produces a refined feature specification that
feeds into the RESEARCH phase. You can proceed directly to /team <spec>
with the output, or run /team-brainstorm first and then /team with the
refined spec.
Ask one to three focused questions about the feature:
Do not ask more than three questions at once. If you need more, ask in rounds after getting answers.
For the stated goal, briefly sketch two to three approaches:
## Approach A: [name]
[One paragraph: what it does, when it is better, what it trades away]
## Approach B: [name]
[One paragraph: what it does, when it is better, what it trades away]
## Approach C: [name] (optional)
[One paragraph: what it does, when it is better, what it trades away]
Present the approaches and ask which direction resonates, or whether there is a different direction not covered.
Once a direction is chosen, negotiate the scope boundary:
The output of scope negotiation is a short list of in-scope requirements that becomes the input to RESEARCH.
When the brainstorming session reaches agreement, output a refined feature specification:
## Refined Feature Specification
**Feature:** [name]
**Problem:** [one sentence: what pain does this solve?]
**Approach:** [which approach was chosen and why]
**Scope:**
- [in-scope requirement 1]
- [in-scope requirement 2]
- ...
**Out of scope:**
- [explicitly excluded item 1]
- [explicitly excluded item 2]
**Success criteria:**
- [how we know it's working]
**Suggested next step:** /team <refined spec>
/team command the user can run.User: /team-brainstorm add a notification system
Agent: To shape this well, a few questions:
1. What triggers a notification — pipeline events (e.g., verification failed),
user-initiated actions, or both?
2. Who receives notifications — only the person who ran /team, or configurable
recipients?
3. Where should notifications go — terminal output only, or external channels
like Slack or email?
User: Pipeline events only, just the person who ran it, terminal output is fine
but Slack would be great too.
Agent: Based on that, here are two approaches:
...
[session continues]
...
## Refined Feature Specification
**Feature:** Pipeline event notifications
**Problem:** Users running /team have no visibility into pipeline progress
without watching the terminal continuously.
**Approach:** Event-log-driven notifications that emit terminal output for
each pipeline phase transition, with optional Slack webhook for async delivery.
**Scope:**
- Emit a terminal notification on each major phase transition
- Optional Slack webhook via environment variable
- Notify on pipeline completion (success and failure)
**Out of scope:**
- Email notifications
- Per-agent granular notifications
- Notification preferences UI
**Success criteria:**
- Running /team with TEAM_SLACK_WEBHOOK set sends a Slack message on completion
- Terminal output shows phase transitions without user intervention
**Suggested next step:** /team add pipeline event notifications with terminal output and optional Slack webhook
data-ai
Todo-first progress convention for multi-step procedures — loaded by every multi-step agent to track its own steps without drift
testing
Adversarially review a technical design document with fresh context before the human gate. Dispatches the built-in `general-purpose` subagent (clean context, no shared history with the design-author) against `docs/plans/<id>/design.md` and presents its verdict — APPROVE, REQUEST CHANGES, or COMMENT. Optional, not part of the QRSPI pipeline. Trigger on "review the design doc", "audit design.md", "is this design ready", or `/eng-design-doc-review`.
development
Generator-evaluator separation and review methodology — loaded by review agents to enforce fresh-context review discipline, Conventional Comments format, and gate verdicts
data-ai
Prepare one or more isolated git worktrees — one per repository the topic touches. Router action — no agent. Trigger on "set up the worktree", "isolate this work", or "/team-worktree".