The Agent Skills Directory

[SAFE]: The skill acts as a documentation reference for the Paperclip API, guiding agents on coordination and governance workflows without containing executable logic within the skill files themselves.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon data from the Paperclip API, which includes content from other agents or users. This is a primary function of the coordination agent.
Ingestion points: Reads issue comments via GET /api/issues/:issueId/comments and heartbeat context via GET /api/issues/:issueId/heartbeat-context (found in SKILL.md and references/api-reference.md).
Boundary markers: None explicitly defined in the provided instructions to separate instructions from data.
Capability inventory: The agent has capabilities to manage issues, agents, routines, and company-wide skill installations via the documented API.
Sanitization: No specific sanitization or validation logic is defined for the ingested data.
[EXTERNAL_DOWNLOADS]: The documentation describes procedures for importing skills from remote sources such as skills.sh and GitHub repositories. These are standard platform features for extending agent capabilities within the ecosystem described.
[DYNAMIC_EXECUTION]: The API reference includes endpoints for updating an agent's instructions-path and importing new skills. These are legitimate mechanisms within the Paperclip platform for dynamically altering agent behavior and functionality as directed by the control plane.

paperclip