agentleFS
tachyon-zcashtachyon-zcash/ragu/.claude/skills/master-fv/SKILL.md

master-fv

Skill61 starsChanged 6 months agoApache-2.0View on GitHub

Run Ragu formalization work through fv-docs, fv-review, spec-fidelity-audit, and fv-preflight in that order, fixing each stage's findings before advancing. Use when implementing or revising a formalization, or when asked for /master-fv or the full FV cycle.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run sequentially, in this exact order:

    line 16
  2. Distinctive:

    [fv-docs](../fv-docs/SKILL.md): Read the authoritative in-tree FV pages relevant to the task.

    line 18
  3. Distinctive:

    [fv-preflight](../fv-preflight/SKILL.md): Run its merge-blocker audit and required final sweep on the resulting changes, including the …

    line 29
  4. Distinctive:

    Keep one findings list across stages.

    line 39
  5. Distinctive:

    If a finding needs unavailable evidence, a dependency, or a design decision, report it as a blocker with what is needed to resolve it.

    line 44

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Master FV

Invoke each of the four maintained FV skills exactly once, in the order below. At each stage, read and apply its linked SKILL.md in full, fix what it flags, and validate those fixes before invoking the next skill. Use the files in the target checkout so refinements to those skills carry into this workflow. Invoking master-fv invokes all four stages, including those that normally require named invocation.

Workflow

Run sequentially, in this exact order:

  1. fv-docs. Read the authoritative in-tree FV pages relevant to the task. Ground the formalization and any prose edits in those definitions, keeping assumptions and coverage limits explicit.
  2. fv-review. Apply its reimplementation, interface, specification, and proof guidance while implementing or reviewing the scoped formalization. Investigate proof failures and newly added hypotheses as possible under-constraint findings.
  3. spec-fidelity-audit. Audit design intent, implementation, and formalization together. Keep its finding tags and per-claim coverage map, including gaps in the proof-to-code connection and unavailable evidence.
  4. fv-preflight. Run its merge-blocker audit and required final sweep on the resulting changes, including the foundational trust, coverage, and non-vacuity checks. Record actual command results and any checks that could not run.

The linked skills own their detailed procedures. Keep that guidance in its existing files rather than duplicating it here.

Finish

Keep one findings list across stages. Resolve each stage's actionable findings within the authorized scope before advancing. Rerun affected validation as needed, keeping the skills in a single ordered pass. Finish after fv-preflight's findings are addressed and its required checks pass.

If a finding needs unavailable evidence, a dependency, or a design decision, report it as a blocker with what is needed to resolve it; complete any independent work without counting the blocked stage as passed.

Briefly summarize fixes and validation results, and note any remaining blockers.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.