edge-computing-security

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFEDATA_EXFILTRATIONPERSISTENCEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION]: The payloads.md file includes a JavaScript snippet for Cloudflare Workers designed to exfiltrate sensitive response data to an external server (https://attacker.com/log). While provided as a testing payload, it demonstrates a functional data exfiltration pattern.
  • [PERSISTENCE]: The SKILL.md file explicitly instructs the user or agent to establish persistence as part of its methodology, specifically mentioning the use of configuration changes, scheduled tasks, and backdoors to maintain access after an initial compromise.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, potentially untrusted sources such as target website configurations, DNS history APIs (e.g., SecurityTrails), and certificate transparency logs (e.g., crt.sh).
  • Ingestion points: External data is retrieved via WebFetch, WebSearch, and curl commands in payloads.md and SKILL.md.
  • Boundary markers: The instructions do not define clear delimiters or "ignore embedded instructions" warnings for the external content being processed.
  • Capability inventory: The skill has access to Bash, Write, and Edit tools, allowing it to execute commands and modify local files based on the processed input.
  • Sanitization: There is no evidence of input validation, escaping, or filtering of the content retrieved from external targets before it is presented to the agent context.
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute a variety of reconnaissance and exploitation tools (e.g., nmap, dig, dnsrecon, curl) against external infrastructure, which is consistent with its stated purpose of security testing.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 04:36 PM