The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: Fetches and executes an installation script for the 'uv' tool from Astral's official domain.
Evidence: curl -LsSf https://astral.sh/uv/install.sh | sh in SKILL.md.
[EXTERNAL_DOWNLOADS]: Directs the user or agent to download binary releases from Gitea's public release page.
Evidence: https://gitea.com/gitea/gitea-mcp/releases in SKILL.md.
[COMMAND_EXECUTION]: Performs privileged system-level operations, including moving external binaries to system-wide binary paths and granting execution permissions.
Evidence: cp gitea-mcp /usr/local/bin/ in SKILL.md.
Evidence: chmod +x /usr/local/bin/gitea-mcp in SKILL.md.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting user-provided configuration strings into shell commands without explicit sanitization.
Ingestion points: User-provided API keys, URLs, and tokens are collected and interpolated into mcporter commands and .env files.
Boundary markers: No delimiters or instructions to ignore embedded commands are present in the interpolation logic.
Capability inventory: Includes file system write access (cat >), permission modification (chmod), and arbitrary command execution (mcporter, sh).
Sanitization: There is no evidence of input validation or shell-escaping for the user-provided data before it is executed or written to files.

mcp-deploy