Skip to content

Add Execution Procedures, split oversized refs, improve structure#1

Open
whiletrue0x wants to merge 4 commits intofeatbit:mainfrom
whiletrue0x:main
Open

Add Execution Procedures, split oversized refs, improve structure#1
whiletrue0x wants to merge 4 commits intofeatbit:mainfrom
whiletrue0x:main

Conversation

@whiletrue0x
Copy link
Copy Markdown

Summary

  • Add ## Execution Procedure with pseudocode to all 17 SKILL.md files — agents now follow structured workflows instead of interpreting prose
  • Add YAML frontmatter (name, description) to all 18 reference files
  • Split oversized reference files: troubleshooting.md (919 lines) → 3 files, environment-variables.md (754 lines) → 2 files
  • Remove redundant sections (When to Use, Source, Read Next) that duplicated frontmatter descriptions across all skills
  • Normalize terminology: consistent "feature flag" across all SDK skills
  • Fix AGENTS.md stale naming example (featbit-dotnet-sdkfeatbit-sdks-dotnet)
  • Improve README: value proposition one-liner, Quick Start table, reference-style badges, wordmark logo
  • Fix graceful-skip conditions and one-sided conditionals in SDK skills
  • Consolidate duplicate code blocks in featbit-sdks-java

Why

Ran a comprehensive skill-forge v3 review and found 5 systemic patterns across all 17 skills. The core issue: every skill was missing an Execution Procedure, causing AI agents to treat skills as passive knowledge documents rather than structured workflows. Full review report included in docs/REVIEW-REPORT.md.

Changes by category

Structure (all 17 skills)

  • Every SKILL.md now has ## Execution Procedure with a pseudocode block
  • Homeless sections removed (~300 lines of redundant context saved)
  • All reference files have YAML frontmatter for tooling support

Oversized files (deployment-docker)

  • troubleshooting.md (919 lines) → troubleshooting-setup.md + troubleshooting-infrastructure.md + troubleshooting-recovery.md
  • environment-variables.md (754 lines) → core vars + environment-variables-otel.md
  • SKILL.md references updated to match new file names

Quality (SDK skills)

  • "feature flag" vs "flag" terminology normalized
  • Vague graceful-skip conditions replaced with explicit triggers
  • One-sided retry conditionals now have failure paths
  • Duplicate code blocks consolidated

README

  • Added wordmark logo (.github/logo-light.svg, .github/logo-dark.svg)
  • Value proposition one-liner replaces technical description
  • Added Quick Start section with trigger examples table
  • Badges converted to reference-style links

Test plan

  • Verify all 17 skills activate correctly with natural language triggers
  • Verify reference file links resolve (no broken paths after splits)
  • Verify npx skills add featbit/featbit-skills still installs correctly
  • Spot-check EP pseudocode matches actual skill behavior

🤖 Generated with Claude Code

whiletrue0x and others added 4 commits March 26, 2026 16:30
Comprehensive quality audit of all 17 skills using skill-forge methodology.
Covers security, structure, quality, and publishing readiness validation.
Result: PASS with 2 minor warnings (nested reference cross-link, collection size at threshold).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Regenerated REVIEW-REPORT.md and REVIEW-REPORT.pdf with fresh skill-forge audit
- Added missing metadata.author: FeatBit to deployment-docker and deployment-aws skills
- All 17 skills now pass full validation (security, structure, quality, publishing)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Full skill-forge review covering all 17 skills and 15 reference files.
Corrects previous report's terminology assessment and adds 3 systemic
patterns + 7 per-skill findings. Also gitignore conversation exports.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Add Execution Procedure with pseudocode to all 17 SKILL.md files
- Add YAML frontmatter to all 18 reference files
- Split troubleshooting.md (919 lines) into 3 files by category
- Split environment-variables.md: extract OTEL config into separate file
- Remove redundant sections (When to Use, Source, Read Next) from all skills
- Normalize terminology: "feature flag" across all SDK skills
- Fix AGENTS.md stale naming example (featbit-dotnet-sdk → featbit-sdks-dotnet)
- Improve README: value proposition, Quick Start, reference-style badges, logo
- Move review report to docs/, remove old PDF report
- Fix graceful-skip conditions and one-sided conditionals in SDK skills
- Consolidate duplicate code blocks in featbit-sdks-java

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant