software-evaluation
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill instructs the agent to read sensitive configuration files, including the '.env' file, during the reconnaissance phase. Accessing files that contain active credentials poses a high risk of data exposure, even in the absence of a network exfiltration step.\n- [INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted code and configuration files, which creates a vulnerability where adversarial content within those files could manipulate the agent's evaluation or behavior.\n
- Ingestion points: The agent reads all files within a project directory, specifically targeting dependency manifests (package.json, go.mod, pyproject.toml, Cargo.toml) and business logic.\n
- Boundary markers: There are no explicit instructions for the agent to treat file content as data only or to ignore instructions found within the code, although the requirement for specific citations acts as a grounded constraint.\n
- Capability inventory: The skill is capable of writing data to local storage, generating Markdown and JSON reports in the 'docs/evaluation/' directory.\n
- Sanitization: The instructions do not define methods for sanitizing, escaping, or filtering content extracted from the analyzed files.
Audit Metadata