Codex

check-traceability

Verify the full refinement chain from use cases through behavioral specs, pseudo-code specs, code, and tests — report coverage at each layer and identify gaps

104 stars

Best use case

check-traceability is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

It is a strong fit for teams already working in Codex.

Verify the full refinement chain from use cases through behavioral specs, pseudo-code specs, code, and tests — report coverage at each layer and identify gaps

Teams using check-traceability should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

  • You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

  • You only need a quick one-off answer and do not need a reusable workflow.
  • You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/check-traceability/SKILL.md --create-dirs "https://raw.githubusercontent.com/jmagly/aiwg/main/.agents/skills/check-traceability/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/check-traceability/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How check-traceability Compares

Feature / Agentcheck-traceabilityStandard Approach
Platform SupportCodexLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

Verify the full refinement chain from use cases through behavioral specs, pseudo-code specs, code, and tests — report coverage at each layer and identify gaps

Which AI agents support this skill?

This skill is designed for Codex.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

Related Guides

SKILL.md Source

# Check Traceability

## Task

Verify the full specification-to-code refinement chain and report coverage at each layer. The traceability chain has 5 layers:

```
Layer 1: Use Cases (UC-NNN)
    └── Layer 2: Behavioral Specs (DES-UCR-NNN, DES-SM-NNN, DES-DT-NNN, DES-MIC-NNN)
         └── Layer 3: Pseudo-Code Specs (DES-PSC-NNN)
              └── Layer 4: Source Code (src/*)
                   └── Layer 5: Tests (test/*)
```

## Natural Language Triggers

Users may say:
- "Check traceability"
- "Are all requirements traced?"
- "Show traceability coverage"
- "What use cases are missing specs?"
- "Traceability report"
- "Are there gaps in the spec chain?"
- "Which UCs don't have realizations?"

## Parameters

### --layer (optional, default: all)
Check a specific layer or all layers:
- `all` — full chain (default)
- `uc` — use cases only (are they well-formed?)
- `behavioral` — UC → behavioral spec coverage
- `pseudo` — behavioral → pseudo-code coverage
- `code` — pseudo-code → source code coverage
- `test` — code → test coverage

### --format (optional, default: table)
Output format: `table` (markdown), `csv`, or `json`.

### --guidance (optional)
Focus traceability on specific areas:
```
--guidance "Focus on authentication module"
--guidance "Only check high-priority use cases"
--guidance "Exclude test layer, just check specs"
```

### --interactive (optional)
Ask questions before starting:
1. Which use cases should we check? (all, by priority, by module)
2. Should we check the full chain or specific layers?
3. Are there known gaps we should flag as expected?

## Execution Steps

### Step 1: Discover Artifacts

Scan `.aiwg/` for artifacts at each layer:

```
Layer 1 — Use Cases:
  Glob: .aiwg/requirements/UC-*.md
  Glob: .aiwg/requirements/use-case-*.md

Layer 2 — Behavioral Specs:
  Glob: .aiwg/requirements/realizations/DES-UCR-*.md
  Glob: .aiwg/architecture/state-machines/DES-SM-*.md
  Glob: .aiwg/architecture/decision-tables/DES-DT-*.md
  Glob: .aiwg/architecture/method-contracts/DES-MIC-*.md
  Glob: .aiwg/architecture/activity-diagrams/DES-ACT-*.md
  Glob: .aiwg/architecture/data-flows/DES-DFS-*.md

Layer 3 — Pseudo-Code Specs:
  Glob: .aiwg/architecture/pseudo-code/DES-PSC-*.md
  Also check: Traceability section of DES-MIC files for PSC references

Layer 4 — Source Code:
  Read pseudo-code spec Traceability sections for "Source Code:" references
  Grep source files for traceability comments (e.g., @spec DES-PSC-NNN)

Layer 5 — Tests:
  Read pseudo-code spec Traceability sections for "Tests:" references
  Grep test files for traceability comments (e.g., @spec DES-PSC-NNN)
```

### Step 2: Build Traceability Matrix

For each use case found, trace forward through all layers:

```
UC-001 (Place Order)
  ├── DES-UCR-001 (realization) ✓
  │    ├── DES-SM-001 (Order state machine) ✓
  │    ├── DES-DT-001 (Discount calculation) ✓
  │    ├── DES-MIC-010 (calculateOrderTotal) ✓
  │    │    ├── DES-PSC-010 (pseudo-code) ✓
  │    │    │    ├── src/order/pricing.ts ✓
  │    │    │    └── test/order/pricing.test.ts ✓
  │    │    └── DES-PSC-011 (pseudo-code) ✗ MISSING
  │    └── DES-MIC-013 (submitOrder) ✓
  │         └── DES-PSC-013 (pseudo-code) ✓
  │              ├── src/order/service.ts ✓
  │              └── test/order/service.test.ts ✗ MISSING
  └── DES-ACT-001 (activity diagram) ✓
```

### Step 3: Calculate Coverage

Report coverage percentage at each layer:

| Layer | Total | Covered | Coverage | Status |
|-------|-------|---------|----------|--------|
| UC → Behavioral | 12 UCs | 10 have realizations | 83% | PASS (≥80%) |
| Behavioral → Pseudo-code | 25 methods | 20 have specs | 80% | PASS (≥80%) |
| Pseudo-code → Code | 20 specs | 18 have source | 90% | PASS |
| Code → Test | 18 modules | 15 have tests | 83% | PASS (≥80%) |

### Step 4: Identify Gaps

For each gap, report:
- **What's missing**: specific artifact ID and type
- **Parent artifact**: what it should trace from
- **Priority**: based on parent UC priority
- **Impact**: what's affected downstream

### Step 5: Generate Report

Output to `.aiwg/reports/traceability-report.md`:

```markdown
# Traceability Report

**Date**: YYYY-MM-DD
**Scope**: All use cases / filtered set
**Overall Status**: PASS | GAPS | FAIL

## Coverage Summary

| Layer Transition | Total | Covered | Gaps | Coverage | Threshold | Status |
|-----------------|-------|---------|------|----------|-----------|--------|
| UC → Behavioral | N | N | N | N% | 80% | PASS/FAIL |
| Behavioral → Pseudo-code | N | N | N | N% | 80% | PASS/FAIL |
| Pseudo-code → Code | N | N | N | N% | — | INFO |
| Code → Test | N | N | N | N% | 80% | PASS/FAIL |

## Gap Details

### Missing Behavioral Specs
| Use Case | Priority | Expected Artifact | Status |
|----------|----------|-------------------|--------|
| UC-NNN | High | DES-UCR-NNN | MISSING |

### Missing Pseudo-Code Specs
| Interface Contract | Method | Expected Artifact | Status |
|-------------------|--------|-------------------|--------|
| DES-MIC-NNN | methodName | DES-PSC-NNN | MISSING |

### Missing Tests
| Source File | Pseudo-Code Spec | Expected Test | Status |
|-------------|-----------------|---------------|--------|
| src/path.ts | DES-PSC-NNN | test/path.test.ts | MISSING |

## Direct-Jump Detection

Use cases that skip specification layers (jump from UC directly to code):
| Use Case | Has Behavioral? | Has Pseudo-code? | Has Code? | Has Tests? | Issue |
|----------|-----------------|------------------|-----------|------------|-------|
| UC-NNN | NO | NO | YES | YES | Skips Layer 2+3 |

## Recommendations
- Priority 1: Create behavioral specs for high-priority UCs missing them
- Priority 2: Create pseudo-code specs for methods in critical paths
- Priority 3: Add tests for untested code files
```

## Backward Compatibility

Projects without the specification-complete layer (no behavioral or pseudo-code specs) continue to work:
- If no `.aiwg/requirements/realizations/` directory exists, skip Layer 2 check
- If no pseudo-code specs exist, skip Layer 3 check
- Report the simpler UC → code → test chain with a note that spec layers are not present
- Never fail a project for missing spec layers if they haven't adopted the specification-complete workflow

## Output

- `.aiwg/reports/traceability-report.md` — full traceability report with coverage and gaps

## References

- @$AIWG_ROOT/agentic/code/addons/aiwg-utils/rules/research-before-decision.md — Read the traceability matrix and all linked artifacts before reporting gaps
- @$AIWG_ROOT/agentic/code/addons/aiwg-utils/rules/human-authorization.md — Report gaps and await owner assignment; do not autonomously close or resolve traceability issues
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/rules/provenance-tracking.md — Traceability requirements and provenance standards this skill enforces
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/skills/security-gate/SKILL.md — Security gate references traceability as a prerequisite criterion
- @$AIWG_ROOT/agentic/code/frameworks/sdlc-complete/flows/gate-criteria-by-phase.md — Gate criteria that reference traceability thresholds

Related Skills

eslint-checker

104
from jmagly/aiwg

Run ESLint for JavaScript/TypeScript code quality and style enforcement. Use for static analysis and auto-fixing.

traceability-check

104
from jmagly/aiwg

Verify bidirectional traceability from requirements to code to tests and identify coverage gaps and orphan artifacts

Codex

regression-check

104
from jmagly/aiwg

Compare current behavior against baseline to detect regressions

Codex

quality-checker

104
from jmagly/aiwg

Validate skill quality, completeness, and adherence to standards. Use before packaging to ensure skill meets quality requirements.

Codex

project-health-check

104
from jmagly/aiwg

Analyze overall project health and metrics

Codex

link-check

104
from jmagly/aiwg

Verify @file references in AIWG skills and agents against the linking contract — per-file or corpus-wide, with optional auto-fix

Codex

flow-handoff-checklist

104
from jmagly/aiwg

Orchestrate handoff validation between SDLC phases and tracks (Discovery→Delivery, Delivery→Ops, phase transitions)

Codex

flow-gate-check

104
from jmagly/aiwg

Orchestrate SDLC phase gate validation with multi-agent review and comprehensive reporting

Codex

citation-check

104
from jmagly/aiwg

Check a file for citation quality and GRADE compliance

Codex

checkpoint

104
from jmagly/aiwg

Create, list, or recover mid-workflow checkpoints so interrupted work resumes from a known-good position

Codex

check-completeness

104
from jmagly/aiwg

Analyze collection completeness against canonical discography and generate prioritized gap report

Codex

aiwg-orchestrate

104
from jmagly/aiwg

Route structured artifact work to AIWG workflows via MCP with zero parent context cost