pseo-orchestrate
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill processes data from discovery and audit phases to determine subsequent implementation steps.
- Ingestion points: Discovery reports, audit findings, and user confirmations processed in SKILL.md.
- Boundary markers: No explicit markers or delimiters are defined to isolate untrusted content from the orchestrator logic.
- Capability inventory: The skill utilizes Bash, Edit, Write, Read, Glob, and Grep to implement SEO changes.
- Sanitization: No explicit sanitization or validation of input data from previous phases is described.
- [Command Execution] (SAFE): The skill uses Bash and file manipulation tools as intended for SEO infrastructure and content implementation. No malicious patterns or privilege escalation attempts were detected.
- [Prompt Injection] (SAFE): Natural instructional language is used to guide the agent through the pipeline; no attempts to bypass safety filters or override system prompts were found.
Audit Metadata