The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from project briefs and architecture documents to perform security analysis. The absence of explicit boundary markers or sanitization logic creates a vulnerability where malicious instructions embedded in these documents could influence the agent's behavior.
Ingestion points: The workflow in SKILL.md identifies ideator-output/project-brief.md, constructor-output/security-flags.md, and user-provided architecture documents as primary input sources.
Boundary markers: There are no instructions for the agent to use delimiters or ignore embedded instructions when reading input files.
Capability inventory: The skill allows the use of write for creating reports and exec for command execution (e.g., mkdir).
Sanitization: No methods for sanitizing or validating the content of ingested files are provided.

datasecurer