Agent skill · operations · daffy0208
Skill Validator
Validates that a skill or MCP implementation matches its manifest by running Codex-powered semantic comparisons across descriptions, preconditions, effects, and API surface.
Why this skill is useful
Adds a comprehensive validation script that ensures skill implementations match their manifests, detecting discrepancies and validating functionality.
What it needs
Requires codex, python3 installed locally. About 5k tokens when loaded. Last updated 2025-12-25. 33 stars on the source repository.
What this skill does
Skill Validator Validate implementations match manifests using Codex for semantic comparison Purpose Ensures that skill/MCP implementations actually deliver what their manifests promise. Uses Codex to perform semantic analysis comparing descriptions, preconditions, and effects against actual code. Detects drift, missing functionality, and over-promised capabilities. When to Use After updating skill implementations During quality audits to verify accuracy When manifests feel outdated or incorrect To detect description-implementation drift Before releasing new versions of resources Key Capabilities Semantic Comparison: Uses Codex to understand if code matches description Precondition Validation: Verifies claimed preconditions are actually checked Effect Verification: Confirms code produces claimed effects API Surface Analysis: Validates exposed functions match manifest Drift Detection: Identifies when implementation diverges from manifest Coverage Scoring: Measures how much of manifest is implemented Inputs Process Step 1: Load Manifest and Implementation Step 2: Validate Description Accuracy Step 3: Validate Preconditions Step 4: Validate Effects Step 5: Validate API Surface Step 6: Generate Validation Report Validation Criteria Scoring Rules Severity Levels High: Missing core functionality, unenforced preconditions, unimplemented effects Medium: Incomplete features, poor error handling, undocumented exports Low: Minor inconsistencies, documentation gaps, style issues Example Output Integration With manifest-generator Validates that generated manifests are accurate by comparing with implementation. With capability-graph-builder Ensures graph relationships are based on accurate capability descriptions. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills daffy0208/skill-validator