genesis-agent-skills-qa
Lint and package Agent Skills with 0 config. Validates skill metadata and specifications.
| Size | Folder | Inside |
|---|---|---|
| 3.6 kB | tests/ | 3 skills · 2 test files |
| 10.8 kB | src/ | 3 source files |
| 4 kB | bin/ | 1 script |
| 6.7 kB | (root) | 5 docs · 1 asset |
| Size | Path |
|---|---|
| 103 B | .gitignore |
| 556 B | CHANGELOG.md |
| 961 B | LICENSE |
| 1.1 kB | QUICKSTART.md |
| 3 kB | README.md |
| 370 B | SECURITY.md |
| 6 B | VERSION · binary |
| 4 kB | bin/genesis-skills-qa.js |
| 641 B | package.json |
| 336 B | src/index.js |
| 2.5 kB | src/packager.js |
| 8 kB | src/validator.js |
| 130 B | tests/fixtures/broken_skill/SKILL.md |
| 249 B | tests/fixtures/dangerous_skill/SKILL.md |
| 369 B | tests/fixtures/valid_skill/SKILL.md |
| 81 B | tests/fixtures/valid_skill/scripts/run_analysis.sh |
| 2.7 kB | tests/test_validator.js |
Already bought this?
View your purchasesREADME
Genesis Agent Skills QA & Packager
Deterministic quality assurance, security auditing, and packaging harness for Agent Skills (SKILL.md) in modern coding-agent ecosystems (Codex CLI, Claude Code, Cursor, Gemini).
1. What Problem Does This Solve?
Teams building custom Agent Skills encounter silent failures in production:
- Malformed YAML frontmatter that agents cannot parse.
- Broken local path references to missing helper scripts.
- Bloated context payloads (>50KB) that blow out token windows.
- Accidental embedding of developer API keys, private keys, or credentials.
- Destructive commands (
rm -rf,sudo,curl | sh) embedded in agent prompts.
genesis-skills-qa deterministically catches these issues in CI or pre-commit before deployment and bundles passing skills into tamper-evident .skill.tar.gz packages with SHA-256 integrity manifests.
2. Who Buys It?
- AI Engineers and Prompt Engineers maintaining Agent Skill libraries.
- Engineering teams deploying autonomous coding agents (Codex / Claude Code / Cursor).
- Enterprise platform teams enforcing security boundaries on agent capabilities.
3. What Does It Output?
- Human-readable terminal diagnostic table with exact line citations.
- Machine-readable JSON output (
--json) for CI/CD gates. - Signed SHA-256 distribution bundle (
.skill.tar.gz) with.skill-manifest.json.
4. How Fast Can You Test It?
Under 5 seconds. Run:
./bin/genesis-skills-qa.js ./tests/fixtures/valid_skill
5. Why Is This Different From Generic Markdown Linters?
Generic markdown linters only check syntax. genesis-skills-qa is purpose-built for the 2026 Agent Skill standard: it audits token context budgets, verifies executable script references, scans for API key entropy, and prevents dangerous escalation commands from entering agent memory.
Installation & Quickstart
# Verify a skill directory
npx genesis-skills-qa ./my-skill
# Run in CI with JSON output and exit code
npx genesis-skills-qa ./my-skill --json
# Package verified skill for distribution
npx genesis-skills-qa ./my-skill --package --out ./dist/my-skill.skill.tar.gz
Diagnostic Rules Evaluated
| Rule ID | Severity | Description |
|---|---|---|
SKILL_MD_EXISTS | FAIL | Checks for presence of SKILL.md at root |
FRONTMATTER_SCHEMA | FAIL | Validates YAML delimiters and key-value structure |
FRONTMATTER_NAME | FAIL | Enforces non-empty name identifier |
FRONTMATTER_DESC | FAIL | Enforces actionable description for agent tool routing |
PAYLOAD_BUDGET | WARN / FAIL | Flags context bloat (>15KB warning, >50KB failure) |
SECURITY_SECRET_LEAK | FAIL | Scans for OpenAI, GitHub, Slack tokens, private keys |
SECURITY_DANGEROUS_CMD | WARN / FAIL | Detects destructive commands (rm -rf, sudo, curl|sh) |
BROKEN_LOCAL_LINK | WARN | Verifies existence of local files referenced in markdown |
License
Commercial — Genesis Code (genesiscode2026). See LICENSE.
Security scan report
Scanned by X402 Git, OSV (opens in a new tab) and Socket (opens in a new tab) — passed on all three. Findings from X402 Git and Socket are listed below.
- secrets
- none found
- X402 Git
- size limits
- within caps
- X402 Git
- known malware
- no dependency manifests to check
- OSV
- known vulnerabilities
- no dependency manifests to check
- OSV
- dependency behaviour
- nothing on a direct dependency · 1 in transitive packages
- Socket
- instruction smuggling
- none found
- X402 Git
- what it can do
- listed below
- X402 Git
What this can do
- Tools
- none
- Runs
- bin/genesis-skills-qa.js
- tests/fixtures/valid_skill/scripts/run_analysis.sh
- Talks to
- nothing
- Reads
- no environment variables
- At install
- nothing runs
Findings (3)
- LOWSocket SBOM Resolver — Socket SBOM Resolver (transitive) — no lockfile, so installs are not reproducible
- LOWbin/genesis-skills-qa.js — Ships the script bin/genesis-skills-qa.js
- LOWtests/fixtures/valid_skill/scripts/run_analysis.sh — Ships the script tests/fixtures/valid_skill/scripts/run_analysis.sh
Releases 1
- v1.0.02026-09-17
Production-ready release of Agent Skills QA & Packager. Validates, packages, and quality-assures agent skill definitions.