Skip to content

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

  • Improved skill quality across 65 skills via Tessl review optimization (#172)
    • Expanded descriptions with capability verbs (what it does + when to use it)
    • Removed redundant “Role Definition” and “When to Use This Skill” sections
    • Added structured workflow validation checkpoints with failure recovery loops
    • Added inline code examples demonstrating key patterns per technology
    • Tightened MUST DO / MUST NOT DO constraints with rationale
    • Removed “Knowledge Reference” keyword lists (absorbed into descriptions and triggers)
  • Narrowed Description Trap from “trigger-only” to “no process steps” — capability verbs now permitted in descriptions per AgentSkills.io spec
  • Updated description format to [Brief capability statement]. Use when [triggering conditions].
  • Validator now checks "Use when" in description instead of startswith("Use when")
  • Standardized README links from HTML to Markdown, improved Quick Start readability (#166)
  • Fixed stale counts on docs site landing page (65 → 66 skills, 357 → 365 references)
  • Added Astro site files to update-docs.py release automation so counts stay in sync
  • @popey — Improve skill quality across 65 skills via Tessl review (#172)
  • @hasan613 — Standardize README links and improve Quick Start section (#166)
  • Issue linking documentation in atlassian-mcp/references/jira-queries.md with parameter semantics, code examples, and anti-patterns (#163)
  • Updated php-pro Symfony console command example to use framework’s built-in exception handling (#164)
  • Added Framework Idiom Principle to CLAUDE.md reference file standards
  • Added inline jira_create_issue_link parameter hints to approve-synthesis and create-implementation-plan commands (#163)
  • Fixed reversed Jira “Blocks” link parameters — inward_issue_key is the blocker, outward_issue_key is the blocked issue (#163)
  • Fixed nested code block rendering in prompt-engineer skill’s CoT example (#160)
  • @wiretail — Report and document reversed Jira issue link semantics (#163)
  • @Big-Shark — Update console example in symfony-patterns (#164)
  • @fiberproduct — Fix nested code block formatting in prompt-patterns.md (#160)
  • Added Agent Skills CLI as alternative installation method in QUICKSTART.md (#151)
  • Upgraded GitHub Actions for Node 24 compatibility (#146)
    • actions/checkout v4 → v6
    • actions/setup-node v4 → v6
    • actions/setup-python v5 → v6
  • Upgraded actions/upload-pages-artifact v3 → v4 (#147)
  • Removed incorrect Python/Pydantic V1 reference from php-pro skill MUST NOT DO section (#154)
  • @salmanmkc — Upgrade GitHub Actions for Node 24 compatibility (#146)
  • @salmanmkc — Upgrade GitHub Actions to latest versions (#147)
  • @Karanjot786 — Add Agent Skills CLI installation method (#151)
  • @Aivanaso — Remove copy-paste error from php-pro skill (#154)
  • Bidirectional cross-reference validation in validate-skills.py (closes #100)
    • CrossRefChecker class: warns when skill A references B but B doesn’t reference A
    • Orphan detection: warns about skills with no incoming or outgoing references
    • New --check crossrefs CLI option for isolated cross-reference validation
    • All issues are WARNING severity (advisory, won’t fail CI)
  • Upgraded Code of Conduct to Contributor Covenant v2.1 (closes #131)
  • Refactored CrossRefChecker._build_graph() to reuse BaseChecker._extract_frontmatter()
  • the-fool — domain-agnostic critical reasoning skill with 5 modes (Socratic questioning, dialectic synthesis, pre-mortem, red team, evidence audit)
    • Two-step AskUserQuestion mode selection for 5 modes within 4-option constraint
    • 6 reference files: mode-selection-guide, socratic-questioning, dialectic-synthesis, pre-mortem-analysis, red-team-adversarial, evidence-audit
    • Added to Workflow category in SKILLS_GUIDE.md
    • Added Critical Thinking decision tree and Decision Validation workflow
  • Fixed sync-content.mjs regex in stripHtmlCommentTags not stripping version markers (dots in 0.4.x didn’t match \w+)
  • Fixed social preview generation command in CLAUDE.md release checklist (added npm install --no-save puppeteer)
  • Added “Mentioned in Awesome Claude Code” badge to README (closes #138)
  • Added Awesome Claude Code stat to social preview
  • Added scripts/validate-markdown.py for detecting markdown parsing errors
    • Checks for HTML comments breaking tables, unclosed code blocks, missing table separators, column count mismatches
    • Integrated into CI workflow and release checklist
  • Fixed markdown bold formatting rendering as raw ** in README.md (#139)
  • Fixed minor syntax error in skill YAML frontmatter (#137)
  • Fixed incorrect JQL syntax in epic planning workflows (Parent ="Epic Link" =)
    • Affected: create-epic-plan, create-epic-discovery, WORKFLOW_COMMANDS.md
  • Fixed markdown parsing errors across skill files:
    • Moved HTML comments above tables in code-reviewer and debugging-wizard skills
    • Added missing closing code fences in incident-chaos, mysql-tuning, postgresql-tuning
    • Escaped pipe character in php-pro/modern-php-features.md table
  • @hesreallyhim — Featured in Awesome Claude Code (#138)
  • @hesreallyhim — Fix formatting and links in README.md (#139)
  • @hesreallyhim — fix: minor syntax error in .md yaml frontmatter (#137)
  • Replaced CONTRIBUTORS.md reference with GitHub contributors link in CONTRIBUTING.md (closes #130)
  • Refactored validate-skills.py with Python 3.11+ best practices (closes #133)
    • Extracted FrontmatterResult dataclass and _extract_frontmatter() helper to eliminate duplicate parsing
    • Created MetadataEnumChecker generic base class for enum field validators
    • Compiled regex patterns at module level for performance
    • Modernized type hints (X | None syntax) and added DFSColor IntEnum
  • update-docs.py now auto-bumps version in ROADMAP “Last updated” line
  • Added __pycache__/ to .gitignore
  • Added AdaL CLI (SylphAI) to supported agents table (#135)
  • @liyin2015 — docs: add AdaL CLI to supported agents (#135)
  • Documentation site built with Astro Starlight (site/) (closes #106)
    • Custom GitHub stars component with live API fetch
    • View as Markdown toggle for skill pages
    • SEO metadata improvements
  • Automated release workflow .github/workflows/release.yml (closes #107)
    • Triggers on version tag push (v*)
    • Validates using reusable workflow
    • Extracts release notes from CHANGELOG.md
    • Builds and deploys docs site to GitHub Pages
    • Creates GitHub Release with extracted notes
  • YAML workflow definitions and DAG manifest for project commands (commands/) (closes #124)
  • related-skills field in skill frontmatter metadata (closes #69)
  • scripts/migrate-frontmatter.py for bulk skill metadata updates
  • Local skill development guide (docs/local_skill_development.md)
  • Restructured primary docs, workflow docs, and skill files
  • Rewrote ROADMAP.md with re-triaged GitHub issues
  • Extended validate-skills.py with related-skills validation
  • Documentation tone consistency improvements (closes #132)
  • Fixed broken links and missing references in cloud-architect skill
  • Added links in SKILLS_GUIDE.md
  • 4 new reference files for vue-expert-js skill (closes #97):
    • composables-patterns.md: Custom composables, ref/reactive patterns, lifecycle hooks
    • component-architecture.md: Props, emits, slots, provide/inject, dynamic components
    • state-management.md: Pinia setup, store patterns, reactive state
    • testing-patterns.md: Vitest/Vue Test Utils, component mounting, mocking
  • version.json for centralized version and count management
  • scripts/update-docs.py for automated documentation updates with marker-based replacement
  • scripts/validate-skills.py for skill YAML and reference file validation
  • HTML comment markers in documentation files for explicit count replacement (fixes #99)
  • AskUserQuestions tool integration in feature-forge skill for structured elicitation during discovery, interview, and validation phases (closes #102)
  • Multi-agent pre-discovery pattern in feature-forge for launching parallel Task subagents with domain-specific skills before the specification workshop (closes #102)
  • discovery-for-feature-forge.md example prompt for multi-agent discovery workflow
  • Reusable GitHub Actions validation workflow .github/workflows/validate.yml (closes #105)
  • CI workflow .github/workflows/ci.yml triggering on PRs and pushes to main (closes #105)
  • CI status badge in README
  • Bloc state management reference for flutter-expert skill (#117)
  • Total reference files: 351 → 356
  • Clarified npx add-skill installation and slash command limitations in README (#96)
  • update-docs.py now uses ... tags instead of broad regex patterns
  • Documentation files (README.md, QUICKSTART.md, ROADMAP.md, social-preview.html) updated with markers
  • feature-forge interview flow rewritten to alternate between structured AskUserQuestions and open-ended elicitation
  • feature-forge constraints updated to require AskUserQuestions for structured choices
  • Minor README tidying
  • Non-standard headers (Reference for:, Load when:) removed from 337 reference files (closes #104)
  • non-standard-headers validation promoted from warning to error in validate-skills.py
  • Pattern matching in update-docs.py no longer incorrectly matches inline comments like # 6 reference files in project structure trees (fixes #99)
  • Missing version comparison link for v0.4.1 in CHANGELOG (#116)
  • @thomassamoul — feat: add bloc state management to flutter expert (#117)
  • @Coopyrightdmin — fix: add missing version link for v0.4.1 (#116)
  • @Chidwan3578 — docs: clarify npx add-skill installation (#96)
  • 50 new reference files for 10 skills:
    • atlassian-mcp (5 files): MCP server setup, Jira queries, Confluence operations, authentication, workflows
    • fine-tuning-expert (5 files): LoRA/PEFT, dataset preparation, hyperparameters, evaluation, deployment
    • ml-pipeline (5 files): Feature engineering, training pipelines, experiment tracking, orchestration, validation
    • pandas-pro (5 files): DataFrame operations, data cleaning, aggregation, merging, performance
    • prompt-engineer (5 files): Prompt patterns, optimization, evaluation frameworks, structured outputs, system prompts
    • rag-architect (5 files): Vector databases, embedding models, chunking strategies, retrieval optimization, evaluation
    • salesforce-developer (5 files): Apex development, LWC, SOQL/SOSL, integrations, deployment
    • shopify-expert (5 files): Liquid templating, Storefront API, app development, checkout, performance
    • spark-engineer (5 files): Spark SQL/DataFrames, RDD operations, partitioning, performance tuning, streaming
    • wordpress-pro (5 files): Theme development, plugin architecture, Gutenberg blocks, hooks/filters, security
  • Release checklist validation step in CLAUDE.md for YAML and reference integrity checks
  • YAML parsing errors in 53 skills caused by unescaped Keywords: pattern in descriptions (fixes #93)
  • Missing reference files for 10 skills that had routing tables but no actual reference files (fixes #92)
  • Total reference files: 304 → 351
  • Skill descriptions now follow trigger-only format consistently
  • New --graph flag for /common-ground command
    • Generates mermaid flowchart diagrams visualizing Claude’s reasoning structure
    • Decision points, chosen paths, alternatives, and uncertainties displayed visually
    • Node colors indicate confidence: green (chosen), yellow (decision point), orange (uncertain), gray (alternative)
  • New references/reasoning-graph.md reference file (280 lines)
    • Mermaid diagram generation patterns and templates
    • Node types, styling rules, and edge label conventions
    • Complete examples with auth system reasoning tree
  • README “Context Engineering” section featuring /common-ground command
  • README Quick Start callout for --graph flag
  • Updated README header stats to include “Context Engineering”
  • Enhanced README Author section with LinkedIn profile links
  • Updated project structure documentation to show commands/common-ground/
  • Total reference files: 301 → 304
  • New /common-ground command for surfacing Claude’s hidden assumptions (#88)
    • Two-phase interactive flow (Surface & Select, Adjust Tiers)
    • Assumption classification by Type (stated/inferred/assumed/uncertain)
    • Confidence tiers (ESTABLISHED/WORKING/OPEN)
    • --list and --check flags for quick access
    • Ground file storage in ~/.claude/common-ground/
    • Introduced hybrid command pattern (COMMAND.md + references/) for complex commands
  • New /approve-synthesis command for reviewing and approving synthesis documents (#87)
    • Reviews proposed tickets from synthesis
    • Resolves blocking decisions before ticket creation
    • Allows add/remove/modify of tickets before Jira creation
    • Creates approved tickets in Jira
  • New vue-expert-js skill for JavaScript-only Vue development (#86)
    • JSDoc typing patterns for Vue components
    • Runtime prop validation
  • Split /synthesize-discovery workflow into two commands (#87)
    • /synthesize-discovery now creates synthesis document with proposed tickets
    • /approve-synthesis handles decision resolution and Jira ticket creation
  • Enhanced vue-expert skill with mobile and build tooling (#86)
    • Quasar framework patterns for mobile-first development
    • Capacitor native plugin integration
    • PWA service workers and offline strategies
    • Vite build configuration and optimization
    • Updated Nuxt reference with Custom SSR + Fastify patterns
  • Total skills: 64 → 65
  • Total reference files: 298 → 301
  • Total project commands: 8 → 9
  • Moved legacy workflow commands from .claude/commands/ to .claude/old-commands/legacy/ to avoid polluting the user’s command namespace when installing the plugin
  • Workflow commands remain available via the plugin’s commands/project/ directory structure for programmatic use
  • 8 project workflow commands organized into 4 phases:
    • Discovery: create-epic-discovery, synthesize-discovery - Research and validate requirements
    • Planning: create-epic-plan, create-implementation-plan - Analyze codebase and create execution plans
    • Execution: execute-ticket, complete-ticket - Implement and complete individual tickets
    • Retrospectives: complete-epic, complete-sprint - Generate reports and close work items
  • New commands/project/ directory structure with organized subfolders
  • Comprehensive workflow documentation (docs/WORKFLOW_COMMANDS.md) with mermaid diagrams
  • Atlassian MCP server setup guide (docs/ATLASSIAN_MCP_SETUP.md)
  • Jira integration for ticket management (read tickets, update status, transitions)
  • Confluence integration for document publishing across all workflow phases
  • Mandatory checkpoint system with user approval gates at each phase
  • 10 new skills bringing total to 64:
    • salesforce-developer, shopify-expert, wordpress-pro, atlassian-mcp
    • pandas-pro, spark-engineer, ml-pipeline, prompt-engineer, rag-architect, fine-tuning-expert
  • Updated plugin.json and marketplace.json with commands field
  • Added project management keywords (jira, confluence, epic-planning, sprint, discovery, retrospectives)
  • Updated README with project workflow commands section and updated project structure
  • Total skills: 54 → 64 (19% increase)
  • Total reference files: 284 → 298
  • 35 new skills converted from agents:
    • Languages (12): python-pro, typescript-pro, javascript-pro, golang-pro, rust-engineer, sql-pro, cpp-pro, swift-expert, kotlin-specialist, csharp-developer, php-pro, java-architect
    • Frameworks (7): nextjs-developer, vue-expert, angular-architect, spring-boot-engineer, laravel-specialist, rails-expert, dotnet-core-expert
    • Infrastructure (5): kubernetes-specialist, terraform-engineer, postgres-pro, cloud-architect, database-optimizer
    • API/Architecture (5): graphql-architect, api-designer, websocket-engineer, microservices-architect, mcp-developer
    • Operations (3): sre-engineer, chaos-engineer, cli-developer
    • Specialized (3): legacy-modernizer, embedded-systems, game-developer
  • 193 new reference files across all new skills
  • Comprehensive language-specific patterns for 12 programming languages
  • Framework-specific best practices for 7 additional frameworks
  • Infrastructure-as-code and cloud architecture patterns
  • Modern API design and microservices architecture patterns
  • test-master: Added QA methodology, automation frameworks, performance testing, security testing
  • code-documenter: Added documentation systems, API documentation patterns, technical writing standards
  • devops-engineer: Added platform engineering, deployment strategies, incident response procedures
  • monitoring-expert: Added performance testing, profiling techniques, capacity planning
  • security-reviewer: Added penetration testing, infrastructure security, compliance scanning
  • fullstack-guardian: Added API design standards, architecture decision records, comprehensive deliverables checklist
  • Total skills: 19 → 54 (184% increase)
  • Total reference files: 91 → 284 (212% increase)
  • Expanded tech stack coverage to 25+ frameworks
  • Added 12 programming languages with deep expertise
  • Enhanced decision trees and skill routing
  • Progressive disclosure architecture for all 19 skills
  • 91 reference files across all skills for contextual loading
  • Routing tables in each SKILL.md pointing to domain-specific references
  • Refactored all SKILL.md files to lean format (~80-100 lines each)
  • 50% token reduction on initial skill load
  • Updated CONTRIBUTING.md with progressive disclosure guidelines
  • Rewrote README.md with architecture documentation and improved structure
  • Optimized all 19 skills with modern patterns
  • Framework version updates (React 19, Pydantic V2, Django 5.0, Flutter 3+)
  • ~42% token efficiency improvements across all skills
  • Standardized frontmatter schema (triggers, role, scope, output-format)
  • Updated plugin.json marketplace schema
  • Set release version for plugin distribution
  • Restructured to correct Claude Code plugin format
  • Fixed plugin directory structure
  • Initial release with 19 comprehensive skills
  • Original Development Skills (6): DevOps Engineer, Feature Forge, Fullstack Guardian, Spec Miner, Test Master, Code Documenter
  • Testing & E2E (1): Playwright Expert
  • Backend Framework Skills (3): NestJS Expert, Django Expert, FastAPI Expert
  • Frontend & Mobile Skills (3): React Expert, React Native Expert, Flutter Expert
  • Workflow Skills (4): Debugging Wizard, Monitoring Expert, Architecture Designer, Code Reviewer
  • Security Skills (2): Secure Code Guardian, Security Reviewer
  • Comprehensive documentation (README, SKILLS_GUIDE, CONTRIBUTING)
  • MIT License
  • Languages: TypeScript, JavaScript, Python, Dart, Go
  • Backend: NestJS, Django, FastAPI, Express
  • Frontend: React, React Native, Flutter
  • Testing: Jest, Playwright, Pytest, React Testing Library
  • DevOps: Docker, Kubernetes, CI/CD
  • Monitoring: Prometheus, Grafana, ELK, DataDog
  • Security: OWASP Top 10, SAST tools