skills/matematicsolutions/awesome-matematic-skills-pl/pierwsze-wrazenie-sedziego-pl/Gen Agent Trust Hub
pierwsze-wrazenie-sedziego-pl
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill operates under a restricted environment, declaring
allowed-tools: [Read]in its frontmatter, which limits the agent's capabilities to read-only operations and prevents unauthorized file system modifications or network access. - [SAFE]: The instructions include robust safety protocols, including a mandatory "Bramka człowieka" (Human Gate) requiring professional oversight and a recommendation for data pseudonymization using the
let-it-beutility before processing sensitive information. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data in the form of legal pleadings (e.g., lawsuits, appeals). While this creates an entry point for potential indirect prompt injection, the risk is effectively mitigated by the skill's restricted toolset (no network/write access) and the strict adherence to a pre-defined, seven-part output template which limits the impact of adversarial text within processed documents.
Audit Metadata