better-proposals-automation

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: References the Rube MCP server endpoint at https://rube.app/mcp for tool integration. This is a known resource associated with the Composio platform and is used for legitimate tool connectivity.
  • [COMMAND_EXECUTION]: Outlines the use of RUBE_MULTI_EXECUTE_TOOL to perform actions on the Better Proposals platform. These commands represent the skill's intended automation logic for managing proposals.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through its integration with external API data.
  • Ingestion points: Data retrieved from the Better Proposals API enters the agent context via tool execution results (SKILL.md).
  • Boundary markers: Absent; there are no specific instructions to delimit or ignore instructions within the retrieved content.
  • Capability inventory: RUBE_MULTI_EXECUTE_TOOL and RUBE_REMOTE_WORKBENCH enable the agent to perform actions on the Better Proposals platform (SKILL.md).
  • Sanitization: Not present; the skill does not define methods for validating or cleaning data retrieved from the remote service.
  • [DATA_EXFILTRATION]: Facilitates the transfer of data to the Better Proposals service for automation purposes. This behavior is transparent and aligned with the primary purpose of the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 02:36 AM