aracli-deploy-management

Pass

Audited by Gen Agent Trust Hub on Mar 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes system-level commands using curl, aws, and jq to manage infrastructure, perform HTTP communication with remote agents, and handle S3 state operations.
  • [EXTERNAL_DOWNLOADS]: Orchestrates the retrieval of binary artifacts (PicoClaw) and system images (Golden Images) from S3 storage to remote fleet agents as part of the deployment workflow.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the processing of untrusted external data.
  • Ingestion points: Reads deployment state snapshots and history from the S3 bucket ara-v1/deploy/ (prd.json and various history log files).
  • Boundary markers: Absent; the skill processes structured JSON data from S3 but does not utilize delimiters or specific instructions to ignore embedded commands or malicious strings within that data.
  • Capability inventory: Possesses significant capabilities including performing curl POST requests to administrative endpoints on remote servers and using the aws CLI to modify state data in S3.
  • Sanitization: No evidence of sanitization or strict schema validation for the data retrieved from S3 before it is displayed to the user or used to inform CLI command parameters.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 20, 2026, 03:29 AM