agentsclimarketplace

Graph traceability agent

Skill Agile-V/agile_v_skills/skills/graph-traceability-agent

Create traceability from Agile V requirements to Understand Anything graph nodes, changed files, and tests, ensuring full evidence chain coverage.From its SKILL.md

Install
npx -y skills add Agile-V/agile_v_skills --skill graph-traceability-agent

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

What its file declares

Copied from the file, not written here

The file declares its own license as CC-BY-SA-4.0. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

5.1 KB, as published. Nobody here has run it

Instructions

Purpose

Create traceability from Agile V requirements to Understand Anything graph nodes, changed files, and tests. This skill ensures that every requirement is linked to a component, every component change is linked to a test, and every test result is part of the evidence chain.


Trigger conditions

Use this skill when:

  • Requirements exist for a change to an existing system.
  • A knowledge graph is available.
  • The evidence bundle needs component-level traceability.
  • A change request modifies existing components.
  • An auditor or reviewer requests a traceability matrix.
  • The Red Team needs to verify predicted vs actual impact.

Inputs

- .agile-v/REQUIREMENTS.md                        required
- .agile-v/phases/01-impact/impact_map.md          required
- .agile-v/phases/01-impact/affected_components.json required
- implementation diff (git diff or patch file)    required
- test results (JSON or JUnit XML)                required
- .agile-v/phases/00-understanding/normalized_graph.json optional but preferred

Outputs

All outputs are written to canonical .agile-v/traceability/.

graph_traceability_matrix.md
req_to_component_links.json
component_to_test_links.json
traceability_gaps.md

Required behavior

Step 1: Link requirements to graph nodes

For each requirement in requirements.md:

  1. Extract the requirement ID and description.
  2. Search affected_components.json for components linked to this requirement.
  3. Search the normalized graph for nodes that match the described behavior.
  4. Record: requirement_id → component_id → path/symbol.

Step 2: Link graph nodes to changed files

For each affected component in affected_components.json:

  1. Confirm whether the component was changed in the implementation diff.
  2. Record change type: modified, added, removed, or not changed.
  3. If a predicted component was not changed, flag it for review.

Step 3: Link requirements to tests

For each requirement:

  1. Find test cases in the test results that cover the requirement.
  2. Use test names, comments, or explicit @covers REQ-XXXX annotations.
  3. Record: requirement_id → test_id → test_path → result.

Step 4: Link changed files to test evidence

For each changed file:

  1. Find tests that import or reference it.
  2. Confirm those tests were executed and their results are in the evidence.
  3. Flag any changed file with no linked test.

Step 5: Identify orphan requirements

An orphan requirement is one with no linked component or no linked test. Record each orphan with the reason (no component found / no test found).

Step 6: Identify orphan changes

An orphan change is a file that was modified in the diff but is not linked to any requirement. Record each orphan change. An orphan change is acceptable if it is justified (e.g., module wiring, formatting fix) but it must be explicitly acknowledged.


Output template: graph_traceability_matrix.md

# Graph Traceability Matrix

| Requirement | Graph Node | File/Symbol | Change Type | Test Evidence | Status |
|---|---|---|---|---|---|
| REQ-0001 | node-042 | src/auth/auth.controller.ts::login | Modified | tests/auth/auth.e2e.ts::login_test | Verified |

## Orphan Requirements

| Requirement | Issue |
|---|---|
| ... | No linked component / No linked test |

## Orphan Changes

| File/Symbol | Issue |
|---|---|
| ... | Changed but not linked to a requirement |

## Traceability Decision

- Pass / Fail / Pass with findings
- Findings: ...

Output schema: req_to_component_links.json

[
  {
    "requirement_id": "REQ-0001",
    "component_id": "node-042",
    "path": "src/auth/auth.controller.ts",
    "symbol": "AuthController.login",
    "confidence": "high"
  }
]

Output schema: component_to_test_links.json

[
  {
    "component_id": "node-042",
    "path": "src/auth/auth.controller.ts",
    "test_id": "TEST-001",
    "test_path": "test/auth/auth.e2e-spec.ts",
    "test_result": "pass|fail|not_run",
    "evidence_path": ".agile-v/evidence/test-results.json"
  }
]

Anti-patterns to avoid

  • Do not invent traceability links that are not supported by evidence.
  • Do not mark orphan changes as acceptable without a justification.
  • Do not skip the traceability gaps document even if there are no gaps.
  • Do not include implementation code in the traceability matrix.

Use typed graph relations from the canonical lifecycle contract: artifacts implements and tests verifies a specific baselined REQ-XXXX revision with baseline evidence; verification evaluates artifacts/test cases. Halt rather than invent links. If AI materially influences outputs at any risk level, update .agile-v/aibom/<task_id>/AI_RUN_MANIFEST.yaml per agile-v-aibom.


Compatible tools

Claude Code, Cursor, Copilot, Codex, Gemini CLI, OpenCode, Cline.

What ships with it: 2 files

2.2 KB alongside SKILL.md

Gives 0 of the 12 instructions most context ai engineering skills give

Counted across 1,328 of the 2,349 authors here whose files we hold, read 2026-09-06

  • Dispatch a fresh subagent for each taskin 76 of 1328, across 59 files
  • Perform spec compliance review before code quality reviewin 44 of 1328, across 34 files
  • Dispatch a final code reviewer after all tasksin 38 of 1328, across 26 files
  • Answer subagent questions before allowing implementationin 36 of 1328, across 26 files
  • Use the least powerful model capable of the taskin 33 of 1328, across 26 files
  • Create a TodoWrite list for all tasksin 32 of 1328, across 22 files
  • Perform a task review after each implementationin 31 of 1328, across 24 files
  • Extract all tasks and context from the planin 29 of 1328, across 20 files
  • Provide full task text to subagentsin 28 of 1328, across 20 files
  • Use git worktrees for isolated workspacesin 25 of 1328, across 20 files
  • Specify the model explicitly when dispatching a subagentin 23 of 1328, across 18 files
  • Execute all tasks from the plan without stoppingin 21 of 1328, across 16 files

Said here and by no other author read

  • Link tests to requirements and risks
  • Create AI run manifest if AI influences outputs
  • link requirements to graph nodes
  • link graph nodes to changed files
  • link changed files to test evidence
  • identify orphan requirements

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once.

Keep looking

Skills are one crate of 613,297. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.