hunt-idor
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill includes numerous shell command templates using curl and ffuf to interact with web APIs. These instructions guide the agent to perform network requests against target domains using various HTTP methods and session tokens to test for authorization failures.\n- [DYNAMIC_EXECUTION]: A Python script snippet is provided for the purpose of creating and writing a numeric wordlist file (ids.txt) to the local filesystem, which is used to automate sequential identifier enumeration.\n- [INDIRECT_PROMPT_INJECTION]: The skill establishes a methodology for the agent to ingest and analyze untrusted data from external sources.\n
- Ingestion points: The agent retrieves and processes JavaScript source code bundles, API responses, and HTML content from target applications via curl.\n
- Boundary markers: No explicit instructions or delimiters are provided to the agent to distinguish between target data and its own task instructions.\n
- Capability inventory: The environment includes network access (curl) and filesystem write access (python).\n
- Sanitization: The skill incorporates specific filtering techniques, such as using grep with regular expressions for UUIDs and numeric IDs, which significantly restricts the agent to processing specific data structures and mitigates the risk of executing embedded adversarial instructions.
Audit Metadata