Repository navigation
v1.0.1: Quality, Packaging & skills.sh Readiness - #1
Merged
Merged
Conversation
8 of 43 skills had unquoted descriptions containing colons which caused YAML strict parser failures in skills.sh CLI (only 35/43 discovered). Quote all 43 descriptions with single quotes for consistency.
- Replace broken 'references/official-sources.md' path in all 43 skills with self-contained reference policies and domain-specific reference files that travel with each skill when installed standalone. - Replace the ~900-byte 'Non-negotiable engineering rules' block (duplicated in all 43 skills) with a compact 'Standalone safety' section containing only the critical standalone invariants. - Create portable reference files in skills that need them: routing-matrix, security-checklist, performance-checklist, nui-contract-pattern, and domain-specific source links. Before: 143,086 bytes total SKILL.md After: 86,875 bytes total SKILL.md Reduction: 56,211 bytes (39.3%) Largest skill: fivem-security at 2,523 bytes Full engineering rules remain in AGENTS.md (the constitution). Each skill retains the 3 most critical standalone invariants.
nuiFetch now catches and normalizes all error cases instead of leaking
exceptions through the Promise<NuiResult<TRes>> contract:
- AbortError (timeout) -> { code: 'timeout' }
- Network failure -> { code: 'network_error' }
- Non-2xx HTTP -> { code: 'http_error' }
- Invalid JSON -> { code: 'invalid_response' }
- Unexpected shape -> { code: 'invalid_response' }
Timeout is always cleared in finally block.
No sensitive implementation details leaked in error messages.
New validator covers: - Skill structure (frontmatter, name matching, valid characters, duplication) - YAML strictness (catches unquoted colon descriptions) - Manifest integrity (1:1 filesystem match, category paths, skill count) - Version integrity (VERSION == manifest version) - Portable references (verifies referenced files exist in skill package) - Python syntax (py_compile all helper scripts) - Eval validation (JSON, required fields, skill existence, no duplicates) - Template checks (expected files, no deprecated lua54) Exits non-zero on any failure with actionable file paths.
- Clean project installation verifies 43 skills installed - Unrelated existing skill preservation test - Reinstall without --force skips correctly - Force mode replaces only matching skills - Portable references verified after installation - Root AGENTS.md not overwritten if present - Windows PowerShell installer smoke test - skills.sh discovery structure verification (allowed to fail gracefully)
- Add --dry-run flag - Add --verify flag for post-install validation - Add --version flag - Output an installation receipt (.agents/fivem-agent-skills.json) - Make output summary clearer - Never destructive to unrelated skills - PowerShell script gains full feature parity with Python script
- Add skills.sh compatibility badge to README - Add Quick Start section with npx skills add commands - Add CONTRIBUTING.md guidelines for new skills and evals - Add SECURITY.md policy for vulnerability reporting - Add CHANGELOG.md covering v1.0.0 and v1.0.1 - Add GitHub issue and PR templates
- Add node_modules, coverage, .env, and logs to .gitignore - Update adapter READMEs to mention skills.sh installation
- Rewrote python scanner scripts to use proper argparse and structural logic - Added optional --format json/text output to all scanners for better agent consumption - Improved heuristics in scan_attack_surface.py and scan_hotpaths.py - Cleaned up output formatting for readability
- Update VERSION and skills-manifest.json to v1.0.1
- Fix 'Reinstall without force skips' to properly check for fivem- prefixed skills instead of stale string 'SKIP existing' - Strengthen 'skills-discovery' job by executing the real Vercel skills CLI locally using npx to verify discoverability
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
Implements the complete v1.0.1 Quality & Packaging Patch, making the repository fully ready for distribution through skills.sh / the Vercel
skillsCLI.Changes
nuiFetchto correctly capture and handle timeouts, network errors, HTTP 500s, and JSON parse failures.scripts/validate_skills.pynow structurally verifies all skills, versions, eval schemas, and python scripts.--dry-run,--verify, and receipt generation toinstall.pyandinstall.ps1.CONTRIBUTING.md,SECURITY.md,CHANGELOG.md, and issue/PR templates.All 43 skills are verified, and regression evals are fully passing.