analysis-report

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill allows the agent to install arbitrary Python packages using the uv add <package> command if a required dependency is missing.\n- [COMMAND_EXECUTION]: The agent is instructed to run local analysis scripts (e.g., uv run main.py) and is permitted to write and execute new Python scripts for data analysis tasks.\n- [REMOTE_CODE_EXECUTION]: The capability to add and then run code from third-party packages installed via uv constitutes a remote code execution vector if malicious packages are specified.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted content from the repository's data and reference directories.\n
  • Ingestion points: Files located in data/ (e.g., CSV files) and references/ (e.g., Markdown specifications or standards) are read and analyzed in the skill's workflow (SKILL.md).\n
  • Boundary markers: The instructions do not define any delimiters or system-level warnings to distinguish between raw data and instructions within the processed files.\n
  • Capability inventory: The agent has the ability to execute shell commands (uv run), install packages (uv add), and write files to the filesystem (docs/analysis-report.md).\n
  • Sanitization: There is no mention of escaping, validation, or filtering of the content read from external files before it is used in the context of the report generation or script execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 11:31 AM
Security Audit — agent-trust-hub — analysis-report