context7-cli
Pass
Audited by Gen Agent Trust Hub on Mar 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the
ctx7package globally from the NPM registry or run it vianpx. This is the primary method for distributing the tool. - [EXTERNAL_DOWNLOADS]: The
ctx7 skills installcommand facilitates downloading markdown-based skill definitions from arbitrary GitHub repositories (using the/owner/repoformat). - [COMMAND_EXECUTION]: The skill relies on executing the
ctx7binary with various subcommands (library,docs,skills,setup,login) to perform its functions, which includes modifying agent configuration files (e.g.,.mcp.json,.cursor/mcp.json). - [DATA_EXFILTRATION]: The
ctx7 skills suggestcommand reads local project manifest files (such aspackage.json,requirements.txt,Cargo.toml, andgo.mod) to recommend relevant skills based on project dependencies. This involves processing local file names and contents. - [PROMPT_INJECTION]: The skill's ability to install third-party instructions (skills) from GitHub repositories into the agent's environment creates a surface for indirect prompt injection, as the agent may subsequently follow instructions contained within those markdown files.
Audit Metadata