langfuse
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFE
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill mentions sensitive environment variables such as
LANGFUSE_PUBLIC_KEYandLANGFUSE_SECRET_KEY. However, it explicitly instructs the agent multiple times (e.g., inSKILL.mdandreferences/cli.md) NOT to ask the user to paste these keys into the chat for security reasons. Instead, it suggests setting them via the shell or.envfiles, which is a standard and safe practice for secret management. - [COMMAND_EXECUTION]: The skill uses
npx langfuse-cliandbunx langfuse-clito interact with the Langfuse API. These are standard command-line tools for the Langfuse platform. The allowed tools in the frontmatter strictly limit these commands to thelangfuse-cliandcurloperations targetinglangfuse.comdomains. - [EXTERNAL_DOWNLOADS]: The skill fetches documentation and indices from
https://langfuse.com/llms.txtandhttps://langfuse.com/docs/*.md. These are official documentation sources for the vendor (Langfuse). Per the trusted vendor rules, resources originating from the skill's own vendor (langfuse.com) are considered safe for the intended functionality. - [REMOTE_CODE_EXECUTION]: While the skill uses
npx/bunx(which can download and execute packages), they are pinned to the specific, legitimate vendor CLI toollangfuse-cli. This is the primary intended mechanism for the skill to function as an API bridge for the user's Langfuse instance.
Audit Metadata