Release v1.2.0 - Automated Testing & Modular Architecture

tag v1.2.0
Tagger: aaddrick aaddrick@gmail.com

Release v1.2.0 - Automated Testing & Modular Architecture
Release v1.2.0 - Automated Testing & Modular Architecture

Enhanced test infrastructure with automated iteration loops, modularized orchestrator, and comprehensive documentation improvements.

Automated Test Iteration System

Added product-manager agent (agents/product-manager.md):

• Orchestrates automated test iteration loops
• Analyzes test results and decides iterate vs. stop
• Tracks historical context across iterations
• Manages autonomous fix-test-review cycles
• Integrates with developer agent for continuous improvement

Added developer agent (agents/developer.md):

• Implements fixes autonomously between test iterations
• Analyzes failed tests and creates targeted fixes
• Works within iteration loop without human intervention
• Coordinates with product-manager for decision points

Cross-execution continuity system:

• Sequential iteration numbering (testing/reports/YYYY-MM-DD_N/)
• Preserves historical test data across runs
• Enables trend analysis and regression detection

Added testing/GUIDANCE.md:

• Central decision-making reference for all agents
• Documents test philosophy and approach
• Reduces need for human intervention during iterations
• Ensures consistent agent behavior

Test Orchestrator Modularization

Refactored into clean package structure:

• test_orchestrator/orchestration.py - Main orchestration logic
• test_orchestrator/agents/ - Agent invocation subpackage
  • product_manager.py, developer.py, test_reviewer.py
• test_orchestrator/execution.py - Test execution logic
• test_orchestrator/reporting.py - Report generation
• test_orchestrator/validation.py - Result validation
• test_orchestrator/scenarios.py - Scenario parsing
• test_orchestrator/config.py - Configuration management
• test_orchestrator/models.py - Data models
• Simplified run-all-tests.py to thin entry point

Benefits:

• Improved maintainability and testability
• Clear separation of concerns
• Easier to extend with new features
• Better code organization

Enhanced Features

Parallel test execution:

• Tests can run concurrently for improved performance
• Automatic test-reviewer agent invocation after runs

Duration tracking:

• Individual test timing
• Group execution timing
• Total suite execution time
• All metrics included in reports

--verbose flag for run-all-tests.py:

• Real-time agent output during execution
• Useful for debugging and monitoring progress

Commit ID tracking:

• Links test results to specific git commits
• Enables tracing of fixes across iterations

Post-fix validation:

• Headless mode clarification in test runner
• Ensures fixes are properly validated

Documentation Improvements

Comprehensive exclusion syntax documentation:

• Added prominent "Exclusion Syntax (Critical!)" sections
• Platform-specific examples (Unix --, PowerShell --%)
• Common mistakes and gotchas clearly documented
• Updated all search skills:
  • gh-search-code
  • gh-search-commits
  • gh-search-issues
  • gh-search-prs
  • gh-search-repos

GitHub-wide scope clarification:

• All search skills updated to emphasize cross-repo searching
• Removed single-repository assumptions
• Added explicit scope indicators

Test scenarios improved:

• Explicit GitHub-wide scope indicators
• Removed user-centric placeholders
• Clearer validation criteria
• Better test expectations

Agent Improvements

Test-reviewer agent rewritten:

• Follows subagent best practices
• Better root cause analysis
• More actionable recommendations
• Questions foundational assumptions

Increased agent autonomy:

• Agents halt only for truly new decisions
• Better decision logic for iteration vs. stop
• More autonomous within defined parameters

Master REPORT.md enhanced:

• Includes full test details in Failed Tests Summary
• Better visibility into test failures

Critical Fixes

Test 9 validation issues:

• Resolved with improved validation criteria
• More accurate test expectations

Test accuracy improvements:

• Incorporated reviewer feedback
• Better validation logic

Iteration logic fixes:

• Fixed directory naming to use sequential numbers
• Improved iteration tracking and reporting

Query syntax flexibility:

• Accepts both query:text and --query text as equivalent
• More flexible validation

Skill loading in headless mode:

• Explicitly requires skill loading during tests
• Ensures gh-cli-search skills are available

Test scope ambiguity:

• Clarified in scenarios and expectations
• Reduced false failures

Documentation Updates

• Testing README comprehensively updated for current architecture
• Code structure refactored for improved readability
• CHANGELOG.md updated and reformatted to match Keep a Changelog style
• All versions (1.0.0, 1.1.0, 1.2.0) now use consistent concise format

Version Bump

• Updated plugin.json to v1.2.0
• Updated marketplace.json to v1.2.0
• Comprehensive CHANGELOG.md entry

See CHANGELOG.md for detailed breakdown of all improvements.

Release v1.1.0 - Test Infrastructure Overhaul

tag v1.1.0
Tagger: aaddrick aaddrick@gmail.com

Release v1.1.0 - Test Infrastructure Overhaul
Release v1.1.0 - Test Infrastructure Overhaul

Major performance and reliability improvements to gh-cli-search testing infrastructure.

Test Infrastructure Redesign

Replaced 3-tier agent hierarchy with Python orchestrator:

• Old: test-orchestrator → test-group-leader → test-validator (167 context loads)
• New: run-all-tests.py (80 context loads)
• Performance: 60s+ timeouts → ~6 seconds per test
• Total execution: ~8 minutes for 80 tests

Added Python test orchestrator (testing/scripts/run-all-tests.py):

• Parses test scenarios from markdown files
• Executes 80 tests sequentially with clean isolation
• Generates 3-level reports (master, group, individual)
• Tracks execution time and pass/fail statistics
• Enhanced command extraction for setup commands
• Validates skill usage (gh search vs gh subcommands)

Added test-reviewer agent (agents/test-reviewer.md):

• Post-test analysis for comprehensive result review
• Identifies failure patterns across all tests
• Performs root cause analysis (skill/test/agent/infrastructure)
• Creates REVIEWER-NOTES.md with prioritized recommendations
• Integrated as optional workflow step 3

Critical Fixes

Permission mode bug in run-single-test.sh:

• Changed from invalid 'acceptAll' to 'bypassPermissions'
• Resolved 100% test failure rate

Test timeout issues:

• Disabled episodic memory searches (major speedup)
• Increased timeout from 60s to 120s
• Added concise output mode

Command extraction improvements:

• Enhanced regex for various command formats
• Added support for setup/diagnostic commands
• Better handling of inline vs code block commands

Test prompt improvements:

• "How do I..." → "What command do I run..."
• More explicit requests for code block output

Results

• Pass rate: 70% → 73.8% (59/80 tests)
• Test speed: 60s+ timeouts → ~6s per test
• Timeout rate: 30% → 0%
• Token efficiency: 167 context loads → 80

Skill Updates

testing-gh-skills:

• Updated to reflect Python orchestrator approach
• Marked as SLASH COMMAND ONLY (never auto-invoke)
• Added test-reviewer integration instructions
• Added performance metrics and history

using-gh-cli-search:

• Enhanced with mandatory directive style
• Stronger language to ensure skill usage

Infrastructure

• Added root .gitignore for Python/IDE/OS files
• Added testing/reports/.gitignore for generated reports
• Removed 5 unused slash commands (kept only /test-gh-skills)
• Removed old agent files (test-orchestrator, test-group-leader, test-validator)

Version Bump

• Updated plugin.json to v1.1.0
• Comprehensive CHANGELOG entry for all changes

See CHANGELOG.md for detailed breakdown of all improvements.

Release v1.0.0 - Initial Public Release

tag v1.0.0
Tagger: aaddrick aaddrick@gmail.com

Release v1.0.0 - Initial Public Release
Release v1.0.0 - Initial Public Release

Initial public release of GitHub CLI Search plugin for Claude Code. Provides comprehensive gh CLI search capabilities through skills, slash commands, and automated testing.

Core Features

Nine comprehensive skills for GitHub CLI search:

• gh-search-code - Search code by extension, language, path, content, size
• gh-search-commits - Search commits by author, date range, hash, message
• gh-search-issues - Search issues by label, state, assignee, author, dates
• gh-search-prs - Search pull requests by status, reviews, CI checks, branches
• gh-search-repos - Search repositories by stars, forks, language, topics, licenses
• gh-cli-setup - Installation and troubleshooting guide
• gh-search - General reference for syntax rules
• testing-gh-skills - Orchestrates test suite via agent hierarchy
• using-gh-cli-search - Introduction skill for session start

Six slash commands for quick invocation:

• /gh-search-code, /gh-search-commits, /gh-search-issues
• /gh-search-prs, /gh-search-repos
• /test-gh-skills - Executes full test suite

Testing Infrastructure

Three-tier hierarchical agent architecture:

• test-orchestrator - Manages overall execution, generates master report
• test-group-leader - Coordinates tests within scenario files
• test-validator - Executes individual tests, validates responses

Comprehensive test coverage:

• 80 test scenarios across 6 scenario files
• Validates syntax, quoting, exclusions, special values
• Platform-specific requirements (Unix/Linux/Mac, PowerShell)
• run-single-test.sh for individual test execution

Critical Syntax Guidance

Platform-specific exclusion flags:

• -- flag requirement for Unix-based systems
• --% requirement for PowerShell environments

Special features:

• Proper quoting rules for multi-word queries
• @me special value for current user
• ISO8601 date format with comparison operators

Plugin Infrastructure

Session-start hook system:

• hooks/hooks.json - Hook configuration
• hooks/session-start.sh - Loads using-gh-cli-search skill

Complete plugin structure:

• .claude-plugin/plugin.json - Plugin manifest
• Documentation with examples and best practices
• Installation support (marketplace, clone, manual, team)

Version

• Plugin version: v1.0.0
• Complete CHANGELOG.md with detailed release notes

See CHANGELOG.md for complete feature breakdown.