skills/jkitchin/skillz/pycse/Gen Agent Trust Hub

pycse

Warn

Audited by Gen Agent Trust Hub on Jul 2, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The HashCache feature uses the Python pickle module to save and load cached computation results. Loading data from pickle files is a known security risk that can allow for arbitrary code execution if the files have been tampered with or are from an untrusted source.
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the pycse Python package via pip. This is a legitimate library developed by the skill's author for scientific computing tasks.
  • [PROMPT_INJECTION]: The skill includes functionality to fetch data from Google Sheets using pycse.utils.read_gsheet(url). Ingesting data from external URLs can lead to indirect prompt injection if the source content contains malicious instructions designed to manipulate agent behavior.
  • Ingestion points: pycse.utils.read_gsheet in SKILL.md
  • Boundary markers: None; the skill does not specify markers to separate data from instructions or ignore embedded commands.
  • Capability inventory: The agent has broad permissions via the allowed-tools: ["*"] configuration, enabling file system and shell access.
  • Sanitization: No sanitization or validation logic is mentioned for data retrieved from external URLs.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 2, 2026, 02:12 PM
Security Audit — agent-trust-hub — pycse