projects-staff

Warn

Audited by Gen Agent Trust Hub on Jul 25, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill constructs a bash command to execute a Python script (staffing-score.py) by interpolating a user-provided <portfolio-slug> directly into the command string. If the user provides a slug containing shell metacharacters (e.g., " ; command ; "), this could result in arbitrary command execution on the host system.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes external consultant and project records to generate recommendations.
  • Ingestion points: Reads consultant and project metadata from the cogni-projects directory, specifically projects-portfolio.json and related entity files.
  • Boundary markers: No specific delimiters or instructions are used to distinguish untrusted data from the skill's own instructions during the rendering of the markdown artifact.
  • Capability inventory: The skill has the ability to execute shell commands (via python3) and perform file system write operations to create artifacts and logs.
  • Sanitization: There is no evidence of content sanitization or validation of the fields within the project and consultant records before they are interpolated into the final markdown report.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 25, 2026, 01:51 PM
Security Audit — agent-trust-hub — projects-staff