| name | perform-feature-add-simulation |
| description | Dry-run the add-feature workflow for a proposed feature — friction report and per-phase readiness score, without writing code. |
| disable-model-invocation | true |
Simulate adding the feature the user names — WITHOUT writing a single line of code.
This is a pre-flight check of both the plan and the knowledge layer.
Phases to walk (score each: smooth / friction / blocked)
- Locate — can MODULE_MAP + FEATURE_MAP/CATALOG identify the target modules
without crawling? Name them.
- Plan — draft the touch list per layer; note Stability of every file you'd
change. Any
frozen or ? file ⇒ blocked pending human approval.
- Verify — which suites would prove it works; do they exist and are commands
confirmed?
- Knowledge update — which ai/ files would need updating after the change?
Output
A friction report in chat (and, if asked, saved to ai/analysis/audit-reports/):
per-phase score with the specific missing knowledge that caused friction, total
estimated context cost, and a go / no-go recommendation. Knowledge gaps found here
are the cheapest bugs you will ever fix — file each as an audit TODO.
