Your AI agent workflow is mid. This fixes that. A structured prompt engineering framework for Claude Code. It lives in your git repo as a set of markdown files, JSON configs, and eval templates that give Claude Code consistent behavior, domain awareness, and quality standards.
Stars
2
Forks
0
Watchers
2
Open Issues
0
Overall repository health assessment
^0.100.09
commits
Upgrade excalidraw skill with full drawing specs, color system, templates, and PNG renderer
b9a721bView on GitHubHonesty pass: remove inflated claims, clarify what ships vs external, add pre-commit eval hook, fix unsubstantiated metrics
90675d7View on GitHubReplace markdown/vector context index with structured JSON — context/index.json as L0 master index with schema, tiered loading, session compression
f5ded5aView on GitHubExpanded Quick Start with 3 setup options: local repo + cloud backup, Claude Code session, and Claude Projects
06df5dbView on GitHubAdd OpenViking context management standard — tiered L0/L1/L2 loading, session compression, cost efficiency
67142a0View on GitHubAdd banner, use case diagram, and upgraded README with badges, mermaid, collapsibles, and comparison tables
25752deView on GitHub