codeguard-security-data
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze external data sources, including source code, entity definitions, and system logs, to identify security vulnerabilities. This creates an indirect prompt injection surface where malicious instructions embedded in the analyzed code could attempt to influence the agent's output.
- Ingestion points: The skill reads project source code, data transfer objects (DTOs), and application logs (SKILL.md, references/operations/evidence-playbook.md).
- Capability inventory: The skill provides security recommendations, generates compliance evidence, and suggests code modifications such as adding encryption annotations.
- Boundary markers: The instructions do not define specific delimiters for analyzed content, relying on the agent's contextual understanding.
- Sanitization: Not applicable, as the skill performs static analysis and does not execute the data it processes.
- [EXTERNAL_DOWNLOADS]: The documentation references an installation command for a related security skill within the same ecosystem.
- Evidence: The skill instructs the user to install
codeguard-security-apiusing the commandnpx skills add full-stack-skills/codeguard-skills --skill codeguard-security-api. This resource originates from the skill's own vendor (full-stack-skills).
Audit Metadata