| name | brain-risk-radar |
| description | Proactively scan for sensitive, stale, contradictory, high-impact, or unsafe-to-act-on notes across health, finance, family, legal/admin, and work contexts. |
Brain Risk Radar
Follow AGENTS.md first.
Goal
Surface risks in the knowledge base that could lead to bad real-world decisions if left unexamined.
Inputs to inspect
.generated/stale-claims.md
.generated/unresolved.md
.generated/indexes/open-questions.md
20-knowledge/health/
20-knowledge/finance/
20-knowledge/family/
20-knowledge/work/
- Relevant recent log entries.
Output
- Highest-risk items.
- Why each risk matters.
- What should not be acted on yet.
- Suggested validation source or person.
- One low-friction risk-reduction action.
Judgment
- Prioritize impact and likelihood over completeness.
- Treat stale health, money, family, and work/client claims carefully.
- Keep risk language calm, concrete, and useful.
- Prefer "verify before acting" over alarm.