axiom
Pass
Audited by Gen Agent Trust Hub on Aug 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill acts as a central entrypoint for Axiom observability workflows, including log investigation and dashboard design. Analysis found no evidence of malicious code, data exfiltration, unauthorized network activity, or obfuscation.
- [PROMPT_INJECTION]: The skill handles external data (Axiom logs and metrics), which establishes an indirect prompt injection surface. However, this is inherent to its observability function, and no actual malicious instructions or bypass attempts were found. Ingestion points: Axiom logs and metrics; Boundary markers: None; Capability inventory: Scripts and references in leaf skills; Sanitization: None.
- [SAFE]: The skill uses absolute file paths to route to archived leaf skills. While environment-specific, these paths point to internal skill directories and do not target sensitive system configurations or user credentials.
Audit Metadata