The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data in the form of manuscript drafts, sections, and reviewer comments.
Ingestion points: The skill ingests user-provided text across various classes identified in SKILL.md, such as manuscript, passage, results-flow, and response.
Boundary markers: The instructions lack explicit delimiters or 'ignore embedded instructions' warnings for the data being processed.
Capability inventory: The workflow coordinates multiple high-capability skills that perform file modifications, structural optimization, and integrity audits.
Sanitization: There is no evidence of sanitization, escaping, or validation to distinguish between manuscript content and potential instructions embedded by an attacker within the text.

paper-workflow