The Agent Skills Directory

[SAFE]: The skill facilitates interaction with a local API management service (PIU) running at http://127.0.0.1:3333. All operations are scoped to the user's local environment and interact with a developer's existing tools.\n- [COMMAND_EXECUTION]: The skill utilizes a local script skills/scripts/piu.ts via the Bun runtime. These commands are used to bridge the agent with the local PIU MCP server and represent the skill's core functionality for API management.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes entity data (requests, models, descriptions) retrieved from the PIU server.\n
Ingestion points: Entity data fetched via get-request, get-model, and search-entities as defined in SKILL.md.\n
Boundary markers: None identified in the provided skill instructions.\n
Capability inventory: Network operations via execute_request, variable management (including secrets) via set-vars, and entity modification across 46 specialized tools.\n
Sanitization: No explicit sanitization or validation of the content retrieved from the local PIU server is mentioned.

piu-mcp