write-frontmatter-description
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The documentation suggests using
uvx skilllint@latestto validate skill files. This process involves fetching and executing the latest version of theskilllintpackage from a public registry.\n- [COMMAND_EXECUTION]: The instructions include specific shell commands (uvx skilllint@latest check <file>) for the user or agent to run to ensure frontmatter compliance.\n- [INDIRECT_PROMPT_INJECTION]: The skill is intended to process and rewrite user-supplied description fields, which represents a potential injection surface for untrusted content.\n - Ingestion points: Untrusted description text provided by the user or read from skill files (SKILL.md) in the
Validatesection.\n - Boundary markers: No explicit delimiters or boundary warnings are provided for the agent to use when analyzing the input descriptions.\n
- Capability inventory: The skill possesses the capability to suggest or trigger shell command execution via the
uvxrunner.\n - Sanitization: There is no evidence of specific sanitization or filtering of the external description content before it is processed by the agent.
Audit Metadata