Final — The F-Pass
Skill2 starsChanged 5 months agoApache-2.0View on GitHub
The feature-wide final audit. Re-runs the five lenses across ALL groups together, then two extra passes — Dreamland (dream perfection, then make it real) and Simulation (realistic scale testing). Runs after every group is accepted.
In this repository22 other files it gives its agents.
- A1 — Design & Master-Spec ConsistencySkill
- A2 — Spec ReviewSkill
- A3 — PolishSkill
- A4 — North Star DirectionSkill
- A5 — Web AuditSkill
- AcceptSkill
- AmendSkill
- BrainstormSkill
- ClarifySkill
- ClosureSkill
- ConstituteSkill
- DesignSkill
- DiscoverSkill
- FeatureSkill
- HelpSkill
- ImplementSkill
- InitSkill
- MicroSkill
- North StarSkill
- ResumeSkill
- SpecifySkill
- StatusSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 11
Gate: every group must be accepted in approvals/master.json before the F-pass begins.
- Very specific: line 39
When all seven stages are clean, run the feature review, then /genesis-accept (the review phase), then /genesis-closure to open the PR to …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Final — The F-Pass
After all groups are accepted, the feature gets one final, feature-wide audit pass. The per-group audits caught drift within a group; the F-pass catches what only shows up when the whole feature is viewed at once — cross-group inconsistency, integration gaps, and quality that no single group owned.
Gate: every group must be accepted in approvals/master.json before the F-pass begins. This is the mandatory next step after the last group-accept — never skip straight to closure.
The F-pass is seven stages, run in order. The same Double Gate physics apply: the deeper stages cannot begin until the consistency stage (F1) is clean.
| Stage | Lens | Scope |
| F1 | Design + master-spec consistency (A1, feature-wide) | all groups |
| F2 | Spec review — FR/AC compliance (A2, feature-wide) | every FR/AC in the spec |
| F3 | Polish (A3, feature-wide) | all code/content |
| F4 | North Star direction (A4, feature-wide) | the whole feature vs the vision |
| F5 | Web audit (A5, feature-wide) | the whole stack |
| F6 | Dreamland | dream perfection, then make it real |
| F7 | Simulation | realistic scale testing |
F1–F5 — the lenses, feature-wide
Run each lens exactly as its per-group skill defines it (/genesis-a1…/genesis-a5), but scoped to the entire feature rather than one group. Findings, reports, the zero-tolerance loop, and conviction are identical — see docs/the-audit-system.md. Reports go to audits/final-f{N}.md. F3 cannot start until F1 is at zero findings.
F6 — Dreamland
Step back from what was built and describe the feature at its dream-perfect best — not constrained by what exists, not by the spec, just: what would make this genuinely excellent? Then take each dream and ask whether it can be made real now. Every realizable dream becomes a finding (fix now) or a tracked design input. Dreamland is where the kit reaches past "meets the spec" toward "is actually great." Record the report to audits/final-f6.md.
F7 — Simulation
Run the feature against realistic scale and stress. Load the group simulation scenarios (simulations/g{N}-scenarios.md) and exercise the feature as a whole: realistic data volumes, concurrent use, mid-stream entry, failure and recovery, the unhappy paths. Each scenario is PASS (the feature handles it) or GAP (it doesn't); every gap becomes a finding. Record the report to audits/final-f7.md.
After the F-pass
When all seven stages are clean, run the feature review, then /genesis-accept (the review phase), then /genesis-closure to open the PR to main.
Natural language triggers
- "run the final audits"
- "f-pass"
- "final feature audit"
- "dreamland" / "simulation"Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

