{"skill":{"slug":"agent-skills-context-engineering","displayName":"Agent Skills Context Engineering","summary":"OpenClaw wrapper for Muratcan Koylan's Agent Skills for Context Engineering. 13 skills covering context optimization, multi-agent patterns, memory systems, t...","description":"---\nname: agent-skills-context-engineering\ndescription: >\n  OpenClaw wrapper for Muratcan Koylan's Agent Skills for Context Engineering.\n  13 skills covering context optimization, multi-agent patterns, memory systems,\n  tool design, and evaluation frameworks for production AI agents.\n  Auto-read triggers proactively load the right sub-skill during normal operation\n  (compaction, multi-agent spawns, failure recovery, tool building, memory setup).\nversion: \"1.2.0\"\n---\n\n# Context Engineering Skill\n\nOpenClaw wrapper for [Agent Skills for Context Engineering](https://github.com/muratcankoylan/Agent-Skills-for-Context-Engineering) by [Muratcan Koylan](https://x.com/koylanai).\n\n## Trigger Phrases\n\nActivate when user mentions:\n- \"optimize context\", \"reduce token costs\", \"context compression\"\n- \"design multi-agent system\", \"implement supervisor pattern\"\n- \"implement memory\", \"build memory system\"\n- \"debug agent failure\", \"diagnose context problems\"\n- \"build agent tools\", \"evaluate agent performance\"\n\n## Available Sub-Skills\n\n| Skill | Purpose |\n|-------|---------|\n| context-fundamentals | Core context concepts and anatomy |\n| context-degradation | Lost-in-middle, poisoning, attention patterns |\n| context-compression | Compaction strategies for long sessions |\n| context-optimization | Token budgeting, caching, masking |\n| multi-agent-patterns | Orchestrator, peer-to-peer, hierarchical |\n| memory-systems | Short-term, long-term, graph-based memory |\n| tool-design | Building effective agent tools |\n| filesystem-context | File-based context management |\n| hosted-agents | Background agents with sandboxed execution |\n| evaluation | Agent performance frameworks |\n| advanced-evaluation | LLM-as-judge techniques |\n| project-development | LLM project methodology |\n| bdi-mental-states | BDI cognitive architecture |\n\n## Post-Install Setup (IMPORTANT — check on first load)\n\nThe first time this skill is loaded, check whether context engineering auto-triggers are wired into the user's always-loaded config (typically AGENTS.md, TOOLS.md, or equivalent).\n\n**How to check:** Search for \"Context Engineering Auto-Triggers\" in the user's workspace config files.\n\n**If NOT found**, prompt the user:\n\n> \"I notice the context engineering auto-triggers aren't set up yet. These make your agent proactively apply context engineering best practices — like reading the right guidance before compaction, multi-agent spawns, or failure recovery. Want me to add them to your config? It's 6 rules, takes 30 seconds.\"\n\n**If the user agrees**, add this section to their always-loaded config file (AGENTS.md, TOOLS.md, or equivalent):\n\n```markdown\n## Context Engineering Auto-Triggers\n\nWhen these conditions are detected, read the corresponding sub-skill before proceeding:\n- Before compaction or context exceeds ~80K tokens → read context-compression\n- When spawning 2+ subagents in one session → read multi-agent-patterns\n- On 3+ task retries or agent loop → read context-degradation\n- When building/designing agent tools → read tool-design\n- When setting up memory/persistence → read memory-systems\n- When reading 5+ files as context → read filesystem-context\n\nLoad from: https://raw.githubusercontent.com/muratcankoylan/Agent-Skills-for-Context-Engineering/main/skills/<sub-skill>/SKILL.md\n```\n\n**If already wired**, skip silently.\n\n## Auto-Read Triggers\n\nThese triggers fire **automatically during normal operation** — no user prompt required. When you detect the condition, read the corresponding sub-skill before proceeding.\n\n| Condition | Sub-Skill | When to Fire |\n|---|---|---|\n| Context approaching compaction or token count exceeds ~80K | `context-compression` | Before compaction fires or at compaction boundary |\n| Spawning 2+ subagents in one session | `multi-agent-patterns` | Before first spawn in a multi-agent sequence |\n| Repeated task failure (3+ retries) or agent stuck in loop | `context-degradation` | Before the next retry attempt |\n| Building, designing, or refactoring agent tools/MCP scripts | `tool-design` | When tool construction begins |\n| Setting up memory, JSONL store, vector DB, or persistence layer | `memory-systems` | When memory architecture work begins |\n| Reading 5+ files as context in one session | `filesystem-context` | Before bulk file-loading begins |\n\n**Loading instruction:**\n\n```\nhttps://raw.githubusercontent.com/muratcankoylan/Agent-Skills-for-Context-Engineering/main/skills/<sub-skill>/SKILL.md\n```\n\nRead the sub-skill's SKILL.md, apply its guidance to the current task, then continue. No need to announce this to the user unless they ask why you paused.\n\n## How to Load a Sub-Skill\n\n**Primary method (no setup required):**\n```\nRead: https://raw.githubusercontent.com/muratcankoylan/Agent-Skills-for-Context-Engineering/main/skills/<sub-skill>/SKILL.md\n```\n\n**Optional (for offline use):**\nClone the submodule locally, then read from `references/context-engineering-skills/skills/<sub-skill>/SKILL.md`\n\n## Attribution\n\nAll credit for the underlying context engineering principles belongs to [Muratcan Koylan](https://x.com/koylanai) and contributors to the [original repository](https://github.com/muratcankoylan/Agent-Skills-for-Context-Engineering).\n","tags":{"latest":"1.2.0"},"stats":{"comments":0,"downloads":1150,"installsAllTime":43,"installsCurrent":6,"stars":1,"versions":4},"createdAt":1772304455914,"updatedAt":1778994015907},"latestVersion":{"version":"1.2.0","createdAt":1772308634188,"changelog":"- Added post-install setup instructions to ensure context engineering auto-triggers are wired into user configuration.\n- On first load, skill now checks config files for auto-trigger rules and offers to add them if missing.\n- Provided ready-to-insert config snippet for seamless auto-trigger integration.\n- No changes to core sub-skills or auto-read trigger logic.","license":null},"metadata":null,"owner":{"handle":"levineam","userId":"s176d9bn1j0ncqa4vm4xx8d3s183pxae","displayName":"levineam","image":"https://avatars.githubusercontent.com/u/1668898?v=4"},"moderation":{"isSuspicious":false,"isMalwareBlocked":false,"verdict":"clean","reasonCodes":["review.llm_review"],"summary":"Review: review.llm_review","engineVersion":"v2.4.24","updatedAt":1779967878192}}