The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and returns raw response bodies from the XCrawl API, which creates a surface for indirect prompt injection if the upstream site content contains malicious instructions.
Ingestion points: SKILL.md specifies that the tool returns raw API response bodies from POST https://run.xcrawl.com/v1/map.
Boundary markers: The instructions do not define specific delimiters or instructions for the agent to ignore potential commands within the API response.
Capability inventory: The skill allows Bash(curl), Bash(node), Read, Write, Edit, and Grep.
Sanitization: No explicit sanitization or filtering of the remote content is described.
[DYNAMIC_EXECUTION]: The skill provides template-based Node.js snippets to be executed via node -e for the purpose of reading local configuration and making API requests.
Evidence: Examples in SKILL.md use node -e to read ~/.xcrawl/config.json and execute fetch calls.
[COMMAND_EXECUTION]: The skill utilizes curl and node as its primary tools to interact with external services and manage configuration data.
Evidence: The allowed-tools section and usage examples explicitly rely on shell-based execution of these binaries.

xcrawl-map