The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the agent to fetch protocol specifications and code references from the official NLWeb GitHub repository. This ensures the integration logic remains aligned with the latest server releases.
[DATA_EXFILTRATION]: Provides examples for communicating with an MCP server via JSON-RPC 2.0 POST requests. These network operations are the intended primary function of the skill for tool invocation and do not involve harvesting sensitive local information.
[PROMPT_INJECTION]: The skill facilitates the ingestion of external natural-language data through the ask tool, which creates an inherent surface for indirect prompt injection.
Ingestion points: Responses from the ask tool via the MCP /mcp endpoint in SKILL.md.
Boundary markers: Not explicitly specified in the implementation guidance provided.
Capability inventory: The skill's environment includes tools such as Bash, Write, and WebFetch which could be targeted by instructions embedded in ingested data.
Sanitization: No specific sanitization or validation steps are outlined for the content returned by the NLWeb sites.

nlweb-mcp-server