The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process runtime telemetry data, which constitutes a potential injection surface.
Ingestion points: SKILL.md and references/error-and-crash-capture.md describe the collection of operational logs, error reports, and stack traces.
Boundary markers: SKILL.md requires an explicit event contract and governance controls (redaction, consent) to separate untrusted telemetry from sensitive data.
Capability inventory: The skill allows for the design and modification of Flutter project configurations and telemetry adapters.
Sanitization: references/logging-tracing-privacy.md enforces a default-deny policy for tokens, passwords, and PII, requiring redaction at the earliest possible boundary.

flutter-observability