Loading source
Pulling the file list, source metadata, and syntax-aware rendering for this listing.
Source from repo
A comprehensive collection of Agent Skills for context engineering, multi-agent architectures, and production agent systems.
Files
Skill
Size
Entrypoint
Format
Open file
Syntax-highlighted preview of this file as included in the skill package.
examples/llm-as-judge-skills/skills/index.md
1# Skills Index23Skills are foundational knowledge modules that inform the design and implementation of agents, tools, and prompts.45## Available Skills67### LLM Evaluator8**Path**: `skills/llm-evaluator/llm-evaluator.md`910Covers LLM-as-a-Judge evaluation methodology including:11- Scoring approaches (direct, pairwise, reference-based)12- Evaluation metrics (classification, correlation)13- Known biases and mitigation strategies14- Implementation patterns1516**Key Takeaways**:17- Use direct scoring for objective evaluations18- Use pairwise comparison for subjective preferences19- Always mitigate position bias20- Prefer classification metrics for interpretability2122### Context Fundamentals23**Path**: `skills/context-fundamentals/context-fundamentals.md`2425Covers context engineering principles including:26- Context window management27- Information hierarchy28- Context types (static, dynamic, ephemeral)29- Relevance filtering3031**Key Takeaways**:32- Structure context by priority33- Be explicit over implicit34- Remove redundancy35- Signal freshness of information3637### Tool Design38**Path**: `skills/tool-design/tool-design.md`3940Covers agent tool design best practices including:41- Single responsibility principle42- Input/output schemas43- Error handling patterns44- AI SDK 6 features (approval, strict mode, examples)4546**Key Takeaways**:47- Clear, validated schemas48- Predictable output structure49- Graceful error handling50- Consider approval for dangerous tools5152## Skill Application Matrix5354| Skill | Agents | Tools | Prompts |55|-------|--------|-------|---------|56| LLM Evaluator | Evaluator | directScore, pairwiseCompare | evaluation/* |57| Context Fundamentals | All | All (context params) | All (context handling) |58| Tool Design | All (tool selection) | All | orchestrator-prompt |5960## Adding New Skills61621. Create skill directory: `skills/<skill-name>/`632. Create main file: `skills/<skill-name>/<skill-name>.md`643. Include:65- Overview and purpose66- Core principles67- Practical patterns68- Implementation examples69- References704. Update this index7172## Skill Development Guidelines7374- Focus on principles that transfer across implementations75- Include concrete examples and patterns76- Reference authoritative sources77- Keep content actionable, not just theoretical78- Update as understanding evolves7980