improve-codebase
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted codebase data, including production code and dependency manifests, which could contain adversarial instructions intended to influence the agent's reporting.
- Ingestion points: The agent reads the user's codebase, including production entry points, directory structures, and dependency files as specified in
SKILL.md. - Boundary markers: The instructions lack explicit delimiting or warnings to the agent to disregard instructions embedded within the code or comments it analyzes.
- Capability inventory: The skill is explicitly scoped to be read-only and is instructed not to generate files, modify code, or enter decision-making/implementation flows.
- Sanitization: There are no mentioned mechanisms for sanitizing, escaping, or filtering the analyzed content before it is processed by the agent.
Audit Metadata