The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a bash script (scripts/context7.sh) to interact with the Context7 REST API. It utilizes curl for network requests and jq for processing results. User-provided search queries and documentation topics are sanitized using URI encoding to prevent command or parameter injection.
[DATA_EXFILTRATION]: The skill makes network requests to https://context7.com to retrieve documentation. It allows for an optional API key to be passed via the CONTEXT7_API_KEY environment variable for authentication, which is a standard credential management practice.
[PROMPT_INJECTION]: The skill fetches external text (API documentation and code examples) from a remote service and provides it to the agent. This introduces a surface for indirect prompt injection:
Ingestion points: documentation content fetched via scripts/context7.sh from the context7.com API.
Boundary markers: Absent (the fetched content is not wrapped in delimiters or accompanied by instructions to ignore embedded commands).
Capability inventory: The skill has access to shell tools (curl, jq) and file reading capabilities.
Sanitization: Inputs used to construct the API requests are URI-encoded, but the content returned from the API is not filtered for potential malicious instructions before being passed to the agent.

context7