observability-and-instrumentation
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill includes code patterns that ingest external data from HTTP headers, creating a potential surface for indirect prompt injection if the data is subsequently processed by an agent.
- Ingestion points: The
SKILL.mdfile demonstrates reading thex-request-idheader from incoming requests. - Boundary markers: The provided example includes defensive logic to reject malformed external identifiers and regenerate them if they do not meet safety criteria.
- Capability inventory: Ingested identifiers are used to initialize child loggers and are propagated across service boundaries via HTTP headers.
- Sanitization: The skill provides an explicit sanitization example using a strict regular expression (
/^[A-Za-z0-9_-]{1,64}$/) to validate the input before use. - [SAFE]: The skill promotes significant security best practices, including a hard rule against logging secrets, tokens, or PII. It also recommends using allowlists for log fields and provides a checklist to verify redaction and bounded label sets in staging environments.
Audit Metadata