auraflow
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructions direct the agent to fetch detailed documentation from
https://www.pixazo.ai/models/auraflow.mdto define operational parameters. - [INDIRECT_PROMPT_INJECTION]: The agent is configured to ingest instructions from an external vendor source. Ingestion points: Remote documentation URL in
SKILL.md. Boundary markers: None present. Capability inventory: Subprocess calls (curl), network requests (requests, fetch), and local filesystem access (chmod, file read/write). Sanitization: None provided. - [COMMAND_EXECUTION]: The skill utilizes shell commands for security configuration (
chmod 600on credentials) and suggests usingnpxto install additional tools from the vendor's repository. - [CREDENTIALS_UNSAFE]: The skill manages authentication tokens by reading from and writing to
~/.pixazo/api-key. This is implemented as a standard local credential management practice.
Audit Metadata