Paths: File paths (shared/ , references/ , ../ln-* ) are relative to skills repo root. If not found at CWD, locate this SKILL.md directory and go up one level for repo root. If shared/ is missing, fetch files via WebFetch from https://raw.githubusercontent.com/levnikolaevich/claude-code-skills/master/skills/{path} .

Risk-Based Value Auditor (L3 Worker)

Type: L3 Worker

Specialized worker calculating Usefulness Score for each test.

Purpose & Scope

• Audit Risk-Based Value (Category 3: Critical Priority)

• Calculate Usefulness Score = Impact x Probability

• Make KEEP/REVIEW/REMOVE decisions

• Calculate compliance score (X/10)

Inputs

MANDATORY READ: Load shared/references/audit_worker_core_contract.md .

Receives contextStore with: tech_stack , testFilesMetadata , codebase_root , output_dir .

Workflow

MANDATORY READ: Load shared/references/two_layer_detection.md for detection methodology.

• Parse Context: Extract tech stack, Impact/Probability matrices, test file list, output_dir from contextStore

• Calculate Scores (Layer 1): For each test: calculate Usefulness Score = Impact x Probability 2b) Context Analysis (Layer 2 -- MANDATORY): Before finalizing REMOVE decisions, ask:

• Is this a regression guard for a known past bug? -> KEEP regardless of Score

• Does this test cover a critical business rule (payment, auth) even if Score<10? -> REVIEW, not REMOVE

• Is this the only test covering an edge case in a critical flow? -> KEEP

• Classify Decisions: KEEP (>=15), REVIEW (10-14), REMOVE (<10)

• Collect Findings: Record each REVIEW/REMOVE decision with severity, location (file:line), effort estimate (S/M/L), recommendation

• Calculate Score: Count violations by severity, calculate compliance score (X/10)

• Write Report: Build full markdown report in memory per shared/templates/audit_worker_report_template.md , write to {output_dir}/ln-633--global.md in single Write call

• Return Summary: Return minimal summary to coordinator (see Output Format)

Usefulness Score Calculation

Formula

Usefulness Score = Business Impact (1-5) x Failure Probability (1-5)

Impact Scoring (1-5)

Score Impact Examples

5 Critical Money loss, security breach, data corruption

4 High Core flow breaks (checkout, login, registration)

3 Medium Feature partially broken, degraded UX

2 Low Minor UX issue, cosmetic bug

1 Trivial Cosmetic issue, no user impact

Probability Scoring (1-5)

Score Probability Indicators

5 Very High Complex algorithm, new technology, many dependencies

4 High Multiple dependencies, concurrency, edge cases

3 Medium Standard CRUD, framework defaults, established patterns

2 Low Simple logic, well-established library, trivial operation

1 Very Low Trivial assignment, framework-generated, impossible to break

Decision Thresholds

Score Range Decision Action

=15 KEEP Test is valuable, maintain it

10-14 REVIEW Consider if E2E already covers this

<10 REMOVE Delete test, not worth maintenance cost. Exception: regression guards for known bugs -> KEEP. Tests covering critical business rules (payment, auth) -> REVIEW

Scoring Examples

Example 1: Payment Processing Test

Test: "processPayment calculates discount correctly" Impact: 5 (Critical -- money calculation) Probability: 4 (High -- complex algorithm, multiple payment gateways) Usefulness Score = 5 x 4 = 20 Decision: KEEP

Example 2: Email Validation Test

Test: "validateEmail returns true for valid email" Impact: 2 (Low -- minor UX issue if broken) Probability: 2 (Low -- simple regex, well-tested library) Usefulness Score = 2 x 2 = 4 Decision: REMOVE (likely already covered by E2E registration test)

Example 3: Login Flow Test

Test: "login with valid credentials returns JWT" Impact: 4 (High -- core flow) Probability: 3 (Medium -- standard auth flow) Usefulness Score = 4 x 3 = 12 Decision: REVIEW (if E2E covers, remove; else keep)

Audit Rules

1. Calculate Score for Each Test

Process:

• Read test file, extract test name/description

• Analyze code under test (CUT)

• Determine Impact (1-5)

• Determine Probability (1-5)

• Calculate Usefulness Score

2. Classify Decisions

KEEP (>=15):

• High-value tests (money, security, data integrity)

• Core flows (checkout, login)

• Complex algorithms

REVIEW (10-14):

• Medium-value tests

• Question: "Is this already covered by E2E?"

• If yes -> REMOVE; if no -> KEEP

REMOVE (<10):

• Low-value tests (cosmetic, trivial)

• Framework/library tests

• Duplicates of E2E tests

3. Identify Patterns

Common low-value tests (<10):

• Testing framework behavior

• Testing trivial getters/setters

• Testing constant values

• Testing type annotations

Scoring Algorithm

MANDATORY READ: Load shared/references/audit_worker_core_contract.md and shared/references/audit_scoring.md .

Severity mapping by Usefulness Score:

• Score <5 -> CRITICAL (test wastes significant maintenance effort)

• Score 5-9 -> HIGH (test likely wasteful)

• Score 10-14 -> MEDIUM (review needed)

• Score >=15 -> no issue (KEEP)

Output Format

MANDATORY READ: Load shared/references/audit_worker_core_contract.md and shared/templates/audit_worker_report_template.md .

Write JSON summary per shared/references/audit_summary_contract.md . In managed mode the caller passes both runId and summaryArtifactPath ; in standalone mode the worker generates its own run-scoped artifact path per shared contract.

Write report to {output_dir}/ln-633--global.md with category: "Risk-Based Value" and checks: usefulness_score, remove_candidates, review_candidates.

Return summary per shared/references/audit_summary_contract.md .

When summaryArtifactPath is absent, write the standalone runtime summary under .hex-skills/runtime-artifacts/runs/{run_id}/evaluation-worker/{worker}--{identifier}.json and optionally echo the same summary in structured output.

Report written: .hex-skills/runtime-artifacts/runs/{run_id}/audit-report/ln-633--global.md Score: X.X/10 | Issues: N (C:N H:N M:N L:N)

Note: Tests with Usefulness Score >=15 (KEEP) are NOT included in findings -- only issues are reported.

Critical Rules

MANDATORY READ: Load shared/references/audit_worker_core_contract.md .

• Do not auto-fix: Report only

• Effort realism: S = <1h, M = 1-4h, L = >4h

• Score objectivity: Base Impact and Probability on code analysis, not assumptions

• KEEP tests not reported: Only REVIEW and REMOVE decisions appear in findings

• Cross-reference E2E: REVIEW decisions depend on whether E2E already covers the scenario

Definition of Done

MANDATORY READ: Load shared/references/audit_worker_core_contract.md .

• contextStore parsed successfully (including output_dir)

• Usefulness Score calculated for each test (Impact x Probability)

• Decisions classified: KEEP (>=15), REVIEW (10-14), REMOVE (<10)

• Findings collected with severity, location, effort, recommendation

• Score calculated using penalty algorithm

• Report written to {output_dir}/ln-633--global.md (atomic single Write call)

• Summary written per contract

Reference Files

• Audit output schema: shared/references/audit_output_schema.md

Version: 3.0.0 Last Updated: 2025-12-23