skills/moshe-ship/mkhlab/sarih/Gen Agent Trust Hub

sarih

Pass

Audited by Gen Agent Trust Hub on Jun 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill does not contain any malicious code, obfuscation, or network requests. It serves as documentation and a command interface for a local binary designed for content moderation.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides commands to process external datasets, which is an ingestion surface for untrusted content.
  • Ingestion points: Commands like sarih scan data.jsonl and sarih clean data.jsonl ingest data from local files (SKILL.md).
  • Boundary markers: No explicit prompt boundary markers or instruction-ignoring wrappers are defined in the command usage examples.
  • Capability inventory: The skill executes the sarih binary to analyze and filter text (SKILL.md).
  • Sanitization: The primary purpose of the tool is to identify and filter (sanitize) toxic content, mitigating the risk of processing malicious data.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 26, 2026, 01:16 PM
Security Audit — agent-trust-hub — sarih