The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, creating a surface for indirect prompt injection where malicious instructions could be embedded in the data being analyzed.
Ingestion points: The skill ingests research paper drafts and reviewer comments as described in SKILL.md and references/peer-review-response.md to perform critiques and plan responses.
Boundary markers: Absent. The instructions do not provide delimiters or specific markers to isolate user-provided text from the agent's internal instructions.
Capability inventory: The skill utilizes tools such as copydesk for drafting content and credo-release for submission checks, and it has the capability to generate outlines and responses based on the analyzed data.
Sanitization: Absent. There is no mention of sanitizing, escaping, or filtering the content of the drafts or reviews to prevent instructions within those documents from being executed by the agent.

credo-paper-plan