context-engineering-collection
muratcankoylan/Agent-Skills-for-Context-Engineering
guides agent context management , which improves: more reliable, less costly systems
Details
- judge’s score
- 5/10
- place
- outside the packs
- publisher
- muratcankoylan
- status
- new
- pinned commit
- 9ff0c99 on GitHub (new tab)
- skill’s last commit
- Oct 12, 2026
- repo stars
- 18,177
- main type
- Building with Claude
Judgement
Root SKILL.md is a navigation map to 18 sub-skills absent from the provided set. Broad trigger, mostly conceptual content, no commands or testable rules. Maintained repository, MIT, not excessively promotional.
reservations
- The 18 referenced sub-skills (skills/*/SKILL.md) are not provided: impossible to judge the actual content.
- CHANGELOG, CLAUDE.md, CONTRIBUTING and AGENTS.md annexes truncated: partial judgment.
- The root SKILL.md describes a multi-topic collection (platform-agnostic), not a tool targeting Claude Code.
- Content is largely theoretical and research-oriented; few verifiable commands or rules for a solo developer.
- Triggering
- 1/2
- Added value
- 1/2
- Precision
- 0/2
- Scope
- 1/2
- Quality
- 2/2
Alerts
no alerts
Neither the static scan nor the judge flagged a risk for this skill.
Original description
A comprehensive collection of Agent Skills for context engineering, harness engineering, multi-agent architectures, and production agent systems. Use when building, optimizing, evaluating, or debugging agent systems that require effective context management and reliable operating loops.
added on Oct 19, 2026 · license MIT