skills-finder
Pass
Audited by Gen Agent Trust Hub on Jul 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is designed to execute shell commands for searching and installing software packages. It utilizes the npx utility to run the skills CLI and a local Python script scripts/audit.py to perform security evaluations.
- [EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch security audit data from the skills.sh API. It also downloads and installs external packages from the NPM registry as part of its core functionality.
- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it ingests and processes audit summaries from external security providers. These summaries are used by the agent to determine the safety of a candidate skill.
- Ingestion points: Audit summaries retrieved from https://skills.sh/api/v1/skills/audit/ by the audit.py script.
- Boundary markers: The skill instructions specify that the agent must present these summaries to the user for review when a skill is not explicitly marked as safe.
- Capability inventory: The agent can execute CLI tools and install remote code packages.
- Sanitization: The audit script performs JSON parsing, but the natural language summaries are processed directly by the model without specific sanitization filters.
Audit Metadata