The Agent Skills Directory

[SAFE]: The skill provides legitimate implementation patterns for LLM observability. Analysis of the code and instructions found no evidence of malicious intent, prompt injection, or data exfiltration.
[DATA_EXPOSURE_AND_EXFILTRATION]: The implementation examples correctly use environment variables (process.env.LANGFUSE_PUBLIC_KEY, process.env.LANGFUSE_SECRET_KEY) for sensitive credentials instead of hardcoding them, adhering to security best practices.
[EXTERNAL_DOWNLOADS]: The skill references well-known and reputable resources, including the official Langfuse documentation and the widely-used p-queue utility. These are standard dependencies for the stated functionality.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a potential attack surface by ingesting external trace metadata and telemetry parameters. However, the risk is inherent to observability tools and no exploitable capability was identified.
Ingestion points: Trace metadata, score names, and dataset item inputs processed in SKILL.md and references/implementation.md.
Boundary markers: None explicitly defined in the code snippets for isolating telemetry data.
Capability inventory: The skill uses Read, Write, and Edit tools as defined in the frontmatter configuration.
Sanitization: Not demonstrated in the provided SDK implementation examples.

langfuse-rate-limits