writing-designs
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHPROMPT_INJECTION
Full Analysis
- PROMPT_INJECTION (HIGH): The skill is highly vulnerable to Indirect Prompt Injection (Category 8).
- Ingestion points: The skill reads local
definition.mdfiles and utilizesWebFetchto ingest external content from engineering blogs and GitHub repositories. - Boundary markers: No delimiters or instructions to ignore embedded commands within the fetched data are present.
- Capability inventory: The skill has the capability to write a
design.mdfile to the local filesystem based on the processed untrusted data. - Sanitization: No evidence of sanitization, escaping, or validation of the external content before it is processed or written to disk.
- DATA_EXFILTRATION (LOW): While the skill is designed to fetch information, the combination of file-read access and network-fetch capability creates a theoretical path for data exfiltration if the agent is successfully injected via a malicious issue definition.
Recommendations
- AI detected serious security threats
Audit Metadata