The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides code templates for processing external data from Intercom API responses and webhook payloads.
Ingestion points: Found in references/implementation.md (the instrumented client proxy) and references/examples.md (the webhook handler function).
Capability inventory: The provided code is limited to observability tasks, including metric recording, distributed tracing, and structured logging.
Boundary markers: None explicitly provided in the snippets, as the code is designed for traditional backend service monitoring rather than direct LLM input processing.
Sanitization: The skill follows best practices by implementing a specific PII (Personally Identifiable Information) redaction layer in the Pino logging configuration to ensure sensitive customer data (email, name, phone) is not logged to disk or external systems.

intercom-observability