agent-context-dotnet
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill is designed to read repository configuration files, including sensitive ones like .env and appsettings.json. It includes specific logic in Phase 1f and Phase 4 to identify committed secrets and advise the user on secure management practices instead of including secret values in the output.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes content from the user's repository which may contain untrusted data. It employs a dedicated validation phase (Phase 5) inspired by the Claimify framework to atomize factual claims and require user confirmation, mitigating the risk of accepting malicious instructions embedded in the source code.\n- [SAFE]: The skill uses local file system tools for discovery and strictly limits its output to Markdown documentation. It explicitly forbids writing application code, installing software packages, or running destructive shell commands.
Audit Metadata