Protocol Verification

Verify epistemic protocol consistency before commit through static checks and expert review.

Purpose

Surface potential issues in protocol definitions without blocking commits. Follow Anthropic philosophy: transparency over enforcement, user agency over automation.

Core Principles

Workflow

Phase 1: Static Checks

Run scripts/static-checks.js against project root. This script executes without loading into context.

node ${SKILL_DIR}/scripts/static-checks.js ${PROJECT_ROOT}

Output format:

{
  "pass": [{ "check": "...", "file": "...", "message": "..." }],
  "fail": [{ "check": "...", "file": "...", "message": "..." }],
  "warn": [{ "check": "...", "file": "...", "message": "..." }]
}

Checks performed: scripts/static-checks.js is the orchestrator. Its head names the layer files it runs and what each layer holds, and each layer file's CHECKS list is that layer's inventory. references/verification.md carries why the non-obvious checks exist and what to do when one fires.

Phase 2: Expert Review

Spawn parallel Task subagents for LLM-based review — one per perspective (Type Theory / Category Theory, Instruction Design, Claude Code Ecosystem). Load each subagent's prompt template from references/review-checklists.md (single source of truth; do not duplicate the templates here — see Review Checklists below). Each template samples the protocols changed in the current diff, falling back to a representative sample when the diff is not protocol-scoped.

All three subagents run in parallel. Collect results before proceeding.

Phase 3: Synthesize Findings

Combine static check results with expert review findings. Categorize by severity:

Identify convergence (all perspectives agree) and divergence (perspectives differ).

Apply Claude Code Ecosystem expert's filtered array to dismiss false positives from other perspectives.

Phase 4: Surface via Gate Interaction

Present findings via gate interaction. Format:

## Verification Results

### Critical (n issues)
- [issue 1]: [location] - [description]
- [issue 2]: [location] - [description]

### Concerns (n issues)
- [issue 1]: [location] - [description]

### Notes (n observations)
- [observation 1]

### Filtered (Claude Code context)
- [filtered issue]: [reason dismissed]

---
Select action:

Options to present:

1. Fix critical issues - Address critical findings before commit
2. Review concerns - Examine each concern individually
3. Proceed anyway - Commit with decision logged
4. Cancel - Abort verification

Phase 5: Handle User Decision

Commit message annotation (if proceeding with issues):

[verify: n critical, m concerns acknowledged]

Severity Reference

Consult references/criteria.md for which channel judges which kind of obligation, the severity definitions, and what review reads for.

Review Checklists

Consult references/review-checklists.md for:

• Type Theory / Category Theory expert prompt template
• Instruction Design expert prompt template
• Claude Code Ecosystem expert prompt template
• Known issues checklist
• Synthesis template

Error Handling

Integration Notes

Standalone Usage

Most common pattern: invoke /verify before /commit command.

Output Examples

Clean State

## Verification Results

All checks passed.

- Static checks: all checks pass (0 fail, 0 warn)
- Type/Category review: No issues
- Instruction Design review: No issues

Ready to commit.

Issues Found

## Verification Results

### Critical (1 issue)
- State machine totality: hyphegesis/skills/conduct/SKILL.md - Undefined transition for a declared input with no arm in the step function

### Concerns (2 issues)
- Categorical terminology: hyphegesis/skills/conduct/SKILL.md - limit/colimit may not match intended semantics
- Directive verb: hyphegesis/skills/conduct/SKILL.md - "Invoke AskUserQuestion" should be "call"

### Notes (1 observation)
- Version: hyphegesis plugin.json version not bumped since last change

---
How to proceed?