cobusgreyling — 9 Open-Source AI Agent Skills
cobusgreyling is the author of 9 open-source AI agent skills and MCP servers spanning mcp-server, uncategorized, agent-tool, codex-skill, with a combined 12,025+ GitHub stars. On AgentSkillsHub each is quality-scored (avg 63/100) and security-graded where a README exists (8 of 9) — 8 verified safe.
Top skills by cobusgreyling
- loop-engineering — Practical patterns, starters & CLI tools for loop engineering with AI coding agents. Design systems that prompt and orchestrate agents (insp (11.8k★ · quality 72/100 · security: safe)
- Jev — Unofficial TypeSafe Jev showcase — System One decisions, not chat. (141★ · quality 64/100 · security: safe)
- llm-wiki — A compounding knowledge base maintained by LLM agents — inspired by Karpathy's LLM Wiki pattern (46★ · quality 70/100 · security: safe)
- outerloop — Own the Outer Loop. Evidence → Verdict → Answerability. Governance layer for agentic engineering. (12★ · quality 67/100 · security: safe)
- memory-engineering — Practical patterns, starters & CLI tools for memory engineering with AI coding agents. Stop re-explaining the repo — design the memory. Get (9★ · quality 70/100 · security: safe)
- agent-skills — Personal collection of Agent Skills for AI coding agents (Claude Code, Gemini CLI, Cursor, Codex, etc.) (7★ · quality 56/100 · security: safe)
- claim-ledger — Scores the delta between what a coding agent said it did and the tool results in the transcript. (2★ · quality 64/100 · security: safe)
- is-the-missing-piece-in-ai-agent-tools — Markdown version of the blog post: Is The Missing Piece in AI Agent Tools? (1★ · quality 45/100 · security: safe)
View cobusgreyling on GitHub · Explore audited agent skills