context7-auto-research
Warn
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructions direct the user to install the skill from the GitHub repository
BenedictKing/context7-auto-research. This repository does not match the naming conventions or infrastructure of the stated author and represents an unverified external source. - [REMOTE_CODE_EXECUTION]: The installation method
npx skills adddownloads and integrates code from the third-party repository into the agent's environment. This creates a risk of remote code execution if the repository contains malicious or compromised scripts. - [PROMPT_INJECTION]: The skill fetches documentation from the Context7 API to provide current context to the agent. This serves as an ingestion point for untrusted data, making the agent vulnerable to indirect prompt injection if the fetched content contains malicious instructions.
- Ingestion points: External data retrieved from the Context7 API via the
context7-auto-researchskill. - Boundary markers: None identified; there are no instructions to the agent to treat the fetched documentation as data only or to ignore embedded commands.
- Capability inventory: The environment in which this skill operates (Claude Code) generally has permissions to read/write files and execute shell commands.
- Sanitization: The skill does not describe any methods for filtering or sanitizing the content returned by the API.
- [REMEDIATION_GUIDANCE]: To mitigate these risks, users should verify the contents of the repository before installation. Developers should implement strict boundary markers (e.g., XML tags or clear delimiters) when passing documentation to the LLM and include explicit instructions to ignore any commands found within the external data.
Audit Metadata