debugger
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external, untrusted data such as bug reports, application logs, and source code. This creates an attack surface where malicious instructions could be embedded in the data to influence the agent's behavior. The skill explicitly mentions that the agent may have access to repository, shell, and test-execution tools, which increases the potential impact of a successful injection.
- Ingestion points: The skill processes a "bug report plus whatever code, logs, and context are supplied" as described in SKILL.md.
- Boundary markers: There are no explicit delimiters or instructions provided within the skill to treat the input data as non-executable text or to ignore embedded instructions.
- Capability inventory: The instructions specify that when available, the agent should use "repo, shell, or test-execution tools" to confirm hypotheses, representing high-privilege tool access.
- Sanitization: There is no mention of sanitization, filtering, or validation of the ingested logs or code snippets.
Audit Metadata