JubaKitiashvili/continuous-learning-v2
skills/continuous-learning-v2/SKILL.md
Auto-generate skills and rules from observed React Native development patterns
$ install --global
skillsauthnpx skillsauth add JubaKitiashvili/everything-react-native-expo continuous-learning-v2Install this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 24, 2026, 8:56 PM1.9s1 file scanned
SKILL.md
- name:
- continuous-learning-v2
- description:
- Auto-generate skills and rules from observed React Native development patterns
Continuous Learning v2
This skill manages the continuous learning pipeline — observing patterns during development sessions and converting them into persistent rules and skills.
Architecture
PostToolUse hook (real-time)
→ `continuous-learning-observer.cjs` (lightweight pattern capture)
→ patterns stored in .claude/memory/observations/
/learn command (manual, comprehensive)
→ `extract-session-patterns.js` (full session analysis)
→ `analyze-patterns.js` (pattern clustering + dedup)
→ skill-generator prompt (create new content)
→ `validate-content.js` (verify new content is valid)
/retrospective command (session end)
→ `evaluate-session.js` (quality metrics + suggestions)
How It Works
Real-Time (Automatic)
The continuous-learning-observer.cjs hook runs on PostToolUse events. It:
- Captures the tool name, file paths, and outcome
- Detects repeated patterns (same fix applied > 3 times)
- Stores observations in
.claude/memory/observations/as JSON - Lightweight — adds < 50ms to each tool call
Manual Analysis (/learn)
When the user runs /learn, the pipeline:
- Reads all observations from the current session
- Clusters them by type (style fix, import pattern, architecture choice)
- Compares against existing rules and skills
- Generates candidates for new content
- Presents candidates for user approval
- Writes approved content to
.claude/rules/or.claude/skills/
Session Evaluation (/retrospective)
At session end, evaluate-session.js:
- Aggregates all metrics (files changed, tests added, build status)
- Evaluates which rules triggered and their usefulness
- Suggests rule calibration (tighten/loosen globs, adjust content)
- Generates a session quality report
Configuration
See config.json for tuning parameters:
observationThreshold: How many times a pattern must repeat before flagging (default: 3)maxObservationsPerSession: Prevent memory bloat (default: 100)autoApprove: If true, auto-approve low-risk content (default: false)contentTypes: What types to generate —["rule", "skill"]
Related Skills
JubaKitiashvili/upgrade-workflow
development
VerifiedTrustedCommunity
Guided version migration for React Native and Expo SDK upgrades
31SKILL.mdUpdated Apr 16, 2026
JubaKitiashvili/tdd-workflow
development
VerifiedTrustedCommunity
Test-driven development workflow for React Native — Jest, React Native Testing Library, and Detox
31SKILL.mdUpdated Apr 16, 2026
JubaKitiashvili/security-review
development
VerifiedTrustedCommunity
Mobile security audit for React Native applications
31SKILL.mdUpdated Apr 16, 2026
JubaKitiashvili/performance-optimization
development
VerifiedTrustedCommunity
Step-by-step performance diagnosis and optimization for React Native apps
31SKILL.mdUpdated Apr 16, 2026