What problem does it solve?
This Skill eliminates inconsistent AI agent implementations that deviate from repository conventions, which causes broken tooling workflows, poor agent discoverability, and misconfigured tool access or model settings.
Core Features & Use Cases
- Standardized Agent Frontmatter: Defines required fields (name, description, tools, model, color, skills) and naming rules to ensure all agents are valid and discoverable across the repository.
- File Operation Permissions: Clarifies pre-authorized edit access for
.claude/ and .opencode/ directories to eliminate approval prompts during agent development and maintenance.
- Agent Documentation Templates: Provides reusable templates for tool usage sections, when-to-use guidance, and standardized reference documentation to improve agent usability and reduce documentation drift.
- Model Selection Guidelines: Includes a decision framework for choosing between sonnet and haiku models based on task complexity to balance performance, cost, and output quality.
- Use Case: When building a new docs-checker agent for your OSE monorepo, use this Skill to ensure it has valid frontmatter, documents its tool usage, includes clear when-to-use guidance to avoid confusion with other checker agents, and selects the sonnet model for complex content validation tasks.
Quick Start
Use the agent-developing-agents skill to build a new convention-compliant docs-checker agent for your OSE-style monorepo.
