define-publication
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests content from several local files including
PRODUCT.md,GLOSSARY.md, and thedocs/publications/directory to define guidelines. - Ingestion points:
PRODUCT.md,GLOSSARY.md,docs/publications/,docs/decisions/README.md. - Boundary markers: The skill does not define specific delimiters or warnings to ignore embedded instructions in these files.
- Capability inventory: The skill is capable of writing markdown files to
docs/publications/based on the processed content. - Sanitization: No explicit sanitization or validation of the input file content is described.
- [COMMAND_EXECUTION]: In the 'Evidence of a finished piece' section, the skill specifies that 'every embedded code block and command runs as written'. This instruction could be interpreted by an agent as a requirement to execute shell commands or scripts found within project documentation to verify their correctness, which poses a risk if those files contain malicious commands.
Audit Metadata