The Agent Skills Directory

[SAFE]: The skill utilizes Model Context Protocol (MCP) tools to implement an iterative development workflow. It provides clear instructions for tool discovery and loading which are standard for the supported runtime environment.
[EXTERNAL_DOWNLOADS]: The skill references setup documentation hosted on the vendor's GitHub repository at github.com/Q00/ouroboros. This is an official resource consistent with the skill's purpose and author identity.
[INDIRECT_PROMPT_INJECTION]: The skill implements a workflow that ingests user-provided data into a code execution and evaluation pipeline, which represents a potential injection surface.
Ingestion points: User-provided initial_context is ingested in SKILL.md to begin the evolutionary loop.
Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are defined for the interpolation of context into tool parameters.
Capability inventory: The skill utilizes the ouroboros_evolve_step tool which executes an "Execute→Evaluate pipeline" based on generated ontology and code (SKILL.md).
Sanitization: No explicit sanitization or filtering logic is described for the external content before it enters the tool execution phase.

evolve