The Agent Skills Directory

[PROMPT_INJECTION]: The skill implements handlers for external webhook events where the data (message or text fields) is intended to be processed by AI agents, creating a surface for indirect prompt injection.
Ingestion points: Untrusted data enters the agent context via request bodies in examples/express/src/index.js, examples/fastapi/main.py, and examples/nextjs/app/webhooks/openclaw/route.ts.
Boundary markers: While references/overview.md states payloads are "wrapped with safety boundaries," the provided code examples do not demonstrate implementation of these boundaries.
Capability inventory: The examples currently only perform logging (console.log, print), but the skill's purpose is to facilitate broader integrations (databases, CI/CD, etc.) based on these inputs.
Sanitization: No explicit sanitization or validation of the message content is performed in the example code before usage.
[EXTERNAL_DOWNLOADS]: The documentation and setup instructions recommend the use of hookdeck-cli via npx for local development. This tool is maintained by the skill author.

openclaw-webhooks