wildix-chats-overview
Pass
Audited by Gen Agent Trust Hub on May 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill connects to vendor-controlled domains (api.x-bees.com and chat.wildix-chat.com) to retrieve authentication tokens and chat messages. These operations are limited to the vendor's own infrastructure.
- [COMMAND_EXECUTION]: Executes a local bash script (scripts/get-chats-overview.sh) which contains an embedded Python script to handle data pagination and formatting. It also uses the Glob tool to verify the presence of required peer skills.
- [PROMPT_INJECTION]: The skill processes untrusted third-party data in the form of chat messages, creating an indirect prompt injection surface. The skill explicitly mitigates this by instructing the agent to treat fetched text as data only and to ignore any embedded directives. Ingestion points: Chat messages fetched from the Stream API via the get-chats-overview.sh script. Boundary markers: Explicit security warning provided in SKILL.md. Capability inventory: Bash, Glob, and Skill invocation capabilities allowed in frontmatter. Sanitization: Content is formatted as JSON for structured processing, though no content-level filtering is applied.
Audit Metadata