establishing-project-context

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
  • [SAFE]: No malicious patterns such as credential theft, obfuscation, or remote code execution were found. The skill's behavior is restricted to managing project documentation and glossary terminology.
  • [NO_CODE]: The skill is composed exclusively of Markdown files and does not package any scripts or binary executables.
  • [PROMPT_INJECTION]: The skill was analyzed for indirect prompt injection potential (Category 8).
  • Ingestion points: Reads project terminology from CONTEXT.md and CONTEXT-MAP.md (defined in SKILL.md).
  • Boundary markers: Absent; the skill relies on the agent to parse free-text markdown definitions.
  • Capability inventory: Limited to reading and writing Markdown files in the project workspace.
  • Sanitization: Absent; the skill does not explicitly sanitize the content of the glossary files.
  • Assessment: The risk is assessed as safe as the skill is strictly focused on domain-specific terminology and explicitly instructs the agent to avoid implementation details or technical configurations.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 06:00 AM
Security Audit — agent-trust-hub — establishing-project-context