Repository navigation
Changelog
All notable changes to this project will be documented in this file.
[0.3.0] - 2026-03-26 -- Cycle 2 Complete (Rounds 11-20)
Added
- Stress tests -- 13 tests validating rule engine with 200 mixed promises, 1000 history entries (write/read/trim/clear/trend), rapid init/check/save cycles (100 iterations, interleaved saves, full pipeline, temp file cleanup)
- Real-world integration tests -- comprehensive tests simulating actual multi-step usage patterns
- Advanced user guide -- promise extraction deep-dive, quality score explanation with worked examples, real-world setup walkthrough, FAQ section
Changed
- Test coverage expanded from 193 to 219 tests across 22 test files
- UX refinements -- improved CLI output formatting, error messages with actionable hints
- Code quality -- Cycle 1 code review cleanup, test reorganization, naming consistency
- Documentation -- verified doc-code synchronization, updated all examples to match actual behavior
Security
- Additional input validation layers and path safety hardening
- Error handling audit across all filesystem and parsing operations
- Graceful degradation on every failure path verified
[0.2.0] - 2026-03-26 -- Cycle 1 Complete (Rounds 1-10)
Added
- Performance benchmarks -- 17 tests validating rule engine (100 promises), trend detection (500 entries), state manager (rapid cycles), history bulk ops
- CI integration -- CI-friendly output formats, exit codes, GitHub Actions compatibility
- Cross-tool integration -- documentation for MCP ecosystem and external tool integration
- Input validation -- path traversal protection, YAML bomb prevention, config sanitization
- Error recovery -- graceful handling of corrupted state, atomic writes, filesystem error recovery
- Troubleshooting guide -- common error solutions and usage examples in README
Changed
- Test coverage expanded from 90 to 193 tests across 21 test files
- Code quality improved with refactored modules, better type safety, cleaner boundaries
- User onboarding improved with better CLI help, defaults, and error messages
Security
- Path traversal prevention in all file operations
- YAML bomb protection with 64KB size limit
- Config sanitization for untrusted inputs
[0.1.0] - 2026-03-26
Added
- Promise Extraction --
PromiseCollectorreads CLAUDE.md, memory files, and specs to identify quality promises via LLM extraction - Rule Engine with 7 built-in check types:
file_exists-- verify files or glob patterns existmin_lines-- verify minimum line countscontent_match-- verify required strings in filesglob_count-- verify minimum file counts matching patternsgit_pattern-- verify git commit countsstructure_match-- verify required project structuretrend_check-- verify metrics are not declining over history
- LLM Evaluation -- two-stage pipeline: fast local rules first, AI evaluation only when needed
- Scoring System -- weighted score computation (0-100) with configurable thresholds (healthy/warning/degraded/critical)
- Trend Detection -- analyzes last 5 scores to detect improving/stable/declining patterns
- Context Preservation -- saves and restores session context via
.drift-guard/context.md - Check History -- persists quality reports as timestamped JSON files with trim and clear operations
- MCP Server with 5 tools:
drift_guard_init-- initialize project and collect promise sourcesdrift_guard_track-- track file metadata snapshotsdrift_guard_check-- run quality checks (Stage 1 rules + Stage 2 LLM)drift_guard_save-- save session context for future restorationdrift_guard_report-- generate session quality report
- CLI with 5 commands:
drift-guard init-- set up .drift-guard/ and inject CLAUDE.md instructionsdrift-guard serve-- start MCP server (stdio transport)drift-guard check-- manual quality check with formatted outputdrift-guard report-- session report with drift analysisdrift-guard promises-- display extracted promises table
- CLAUDE.md Auto-Injection --
initcommand appends transparent monitoring instructions - Configurable via
.drift-guard/config.yaml(thresholds, check interval, promise sources, ignore patterns) - Full TypeScript with strict mode, ES2022 target, NodeNext modules
- 90 tests across 15 test files covering all modules and integration scenarios