idx-broker
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the
@membranehq/clipackage from the public NPM registry. This is the official command-line tool for the Membrane platform and is necessary for the skill's functionality. - [COMMAND_EXECUTION]: The skill provides various shell commands using the
membraneCLI to authenticate users, manage connections, and execute API actions. These commands are standard for the described integration and rely on the Membrane platform for secure credential handling. - [INDIRECT_PROMPT_INJECTION]: The skill integrates with external property and lead data from IDX Broker. Like any skill that processes external content, there is a theoretical surface for indirect prompt injection if the ingested data contains malicious instructions.
- Ingestion points: Data retrieved from IDX Broker endpoints (Featured Properties, Leads, Supplemental Listings, etc.) via
membrane action runormembrane requestcommands. - Boundary markers: None explicitly defined in the skill instructions; the agent should rely on internal safety guardrails when processing external data.
- Capability inventory: Execution of shell commands via the
membraneCLI tool, including logging in, listing/running actions, and making proxied HTTP requests. - Sanitization: Not specified within the skill instructions; it is recommended that developers validate or sanitize data retrieved from external APIs before further processing.
Audit Metadata