| name | metacog-preflight |
| description | Classify the task, identify risk, choose a strategy, and set a proof plan before non-trivial work. Use for architecture, migrations, auth, billing, concurrency, large refactors, or whenever failure would be costly. |
| user-invocable | true |
| argument-hint | [task description] |
| allowed-tools | Read, Grep, Glob, Bash |
Metacognitive Preflight
Before acting, produce a compact preflight for $ARGUMENTS:
-
Task type
- bugfix | feature | refactor | architecture | security | research | release / deployment
-
Risk level
- low: cosmetic or local, easy rollback
- medium: repo-wide impact, moderate rollback effort
- high: user-visible, security, data, or production impact
-
Failure cost
- cosmetic
- local breakage
- repo-wide regression
- user-visible incident
- security / data loss
-
Strategy (pick one)
- inspect -> patch -> verify (low risk, known territory)
- map -> plan -> patch -> verify (medium risk, unfamiliar code)
- challenge -> verify -> decide (high confidence needs testing)
- threat-model -> verify -> escalate (security, auth, production)
-
Proof plan
Name the smallest checks that would prove or falsify success (tests, grep, type check, build, manual inspection).
-
Tool plan
Explicitly state which tools are truly necessary and which can be skipped.
Output format
Risk: {level}
Strategy: {chosen}
Proof plan: {checks}
First action: {next step}