coverage-report-analyzer
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill handles coverage reports which are external data sources that could be manipulated to include malicious instructions or adversarial content.\n
- Ingestion points: The agent processes data related to coverage reports and test automation tasks as described in the skill definition (SKILL.md).\n
- Boundary markers: The instructions do not provide explicit delimiters or "ignore instructions" warnings to prevent the agent from following directives potentially embedded within analyzed reports.\n
- Capability inventory: The configuration allows for comprehensive file system and command-line access through Bash(cmd:*), Read, Write, and Edit, creating a vector for actions triggered by malicious inputs (SKILL.md).\n
- Sanitization: There are no logic patterns or instructions provided for validating or sanitizing report content before the agent generates code or configurations based on them.
Audit Metadata