agentskills
Specification and documentation for Agent Skills
Learn to test and evaluate AI agent skill effectiveness before distribution. Our guide uses a zero-dependency validation pipeline to ensure structural correctness, safety, and integrity.
How to Add AI Agent Skill Support to an Existing ClientLearn how to add AI agent skill support to your existing client. Implement a three stage progressive disclosure flow with the skills-ref library to integrate LLM system prompts.
Implementing AI Agent Skill Autocomplete for User-Explicit ActivationLearn how to implement AI agent skill autocomplete by scanning files, parsing front-matter, and exposing skill names for model context and user widgets.
How to Implement Trust Checks for Loading Project-Level AI Agent Skills from Untrusted RepositoriesSecurely load AI agent skills from untrusted repositories with a five-layer validation pipeline in agentskills. Protect your project from malicious code before execution.
Managing AI Agent Skill Context Deduplication to Avoid Duplicate InjectionsLearn how to avoid AI agent skill context deduplication and prevent duplicate injections. The Agent Skills library efficiently manages metadata for consistent performance.
Writing Effective AI Agent Skill Descriptions for Model IdentificationLearn to write effective AI agent skill descriptions for accurate model identification. Craft concise, imperative phrases that AI models use to choose the right skill.
How to Filter and Exclude AI Agent Skills from Catalog Based on User PermissionsLearn how to filter and exclude AI agent skills from your catalog based on user permissions. Securely manage access by validating tool permissions and metadata for a tailored experience.
Implementing Subagent Delegation for Complex AI Agent Skill WorkflowsMaster subagent delegation for complex AI agent workflows with Agent Skills. Learn to extend metadata for hierarchical AI and streamline skill execution. Get started today!
Handling Malformed YAML in AI Agent Skill Frontmatter with Fallback ParsingLearn to handle malformed YAML in AI agent skill frontmatter using strict validation and a fallback parser. Prevent build failures and ensure data integrity.
Creating Cross-Client Compatible AI Agent Skills Using the `.agents/skills/` ConventionLearn to create cross-client compatible AI agent skills using the .agents/skills/ convention. Build portable, domain-specific extensions that work with any compliant client.
Implementing AI Agent Skill Activation: Tools vs. File-Read ApproachesChoose AI agent skill activation: explore tool-based vs file-read methods. Learn how allowed tools and YAML frontmatter enable efficient skill management.
Protecting AI Agent Skill Content from Context Truncation: A Complete GuidePrevent AI agent skill context truncation. Learn how Agent Skills library enforces size limits and uses compact XML summaries to keep full documentation out of LLM prompts.
Have a question about this repo?
These articles cover the highlights, but your codebase questions are specific. Give your agent direct access to the source. Share this with your agent to get started:
curl -s "https://instagit.com/install.md" 
Maintain an open-source project? Get it listed too →