skills/ruvnet/ruflo/cost-export/Gen Agent Trust Hub

cost-export

Pass

Audited by Gen Agent Trust Hub on May 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute a Node.js script located at plugins/ruflo-cost-tracker/scripts/export.mjs. This script is responsible for reading the database and formatting the output.
  • [DATA_EXFILTRATION]: The skill provides functionality to send session telemetry and budget data to external webhook URLs provided via the --webhook argument. While this is the intended purpose of the export tool, it involves transmitting internal agent data to external network endpoints.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests and processes telemetry records from the agent's database.
  • Ingestion points: The script reads session-* and budget-config-* records from the cost-tracking namespace in AgentDB.
  • Boundary markers: No delimiters or explicit instructions to ignore embedded content within the records are present in the skill instructions.
  • Capability inventory: The skill utilizes Bash to execute Node.js scripts and can write to the local filesystem or make network requests.
  • Sanitization: Sanitization is not defined in the skill instructions and is presumably managed within the logic of the external JavaScript exporter.
Audit Metadata
Risk Level
SAFE
Analyzed
May 5, 2026, 04:24 AM
Security Audit — agent-trust-hub — cost-export