agentic-jujutsu
Quantum-resistant, self-learning version control for AI agents with ReasoningBank intelligence and multi-agent coordination
49
28%
Does it follow best practices?
Impact
84%
2.62xAverage score across 3 eval scenarios
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./.claude/skills/agentic-jujutsu/SKILL.mdEvaluation results
95%
75%Adaptive Build Workflow with Learning
Trajectory-based learning
Correct package import
0%
100%
startTrajectory called
0%
100%
Meaningful descriptions
50%
75%
addToTrajectory called
0%
70%
finalizeTrajectory called
0%
100%
Honest score variation
66%
100%
Score within range
100%
100%
getLearningStats called
0%
100%
getPatterns called
0%
100%
JSON.parse on all method results
0%
100%
Failure critique recorded
0%
100%
93%
68%Parallel Code Review Agent System
Multi-agent coordination
Correct package import
0%
100%
Independent JjWrapper per agent
0%
100%
Concurrent execution
100%
100%
No sequential locking
100%
100%
Trajectory per agent
0%
100%
queryTrajectories called
0%
100%
getPatterns called
0%
100%
JSON.parse on method results
0%
100%
Failure critique in finalizeTrajectory
0%
100%
getSuggestion called
0%
0%
64%
13%Artifact Integrity Verification System
Quantum security and operation stats
Correct package for fingerprinting
0%
0%
generateQuantumFingerprint called
0%
50%
verifyQuantumFingerprint called
50%
50%
Fingerprint as hex string
0%
100%
enableEncryption called
100%
100%
Key from crypto.randomBytes
100%
100%
getUserOperations called
100%
100%
getStats called with JSON.parse
0%
0%
stats fields accessed
50%
50%
Trajectory lifecycle complete
100%
100%
- Repository
- ruvnet/ruvector
- Commit
462536e
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.