katana-crawl

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill facilitates the use of local security tools, including katana, subfinder, httpx, and nuclei. These are standard command-line utilities for web reconnaissance and vulnerability scanning.
  • [EXTERNAL_DOWNLOADS]: The skill's primary function is to perform web crawling and URL discovery, which involves making numerous network requests to external, user-specified target domains.
  • [INDIRECT_PROMPT_INJECTION]: The skill interacts with untrusted external content by crawling web pages and extracting data from them.
  • Ingestion points: Data enters the agent's workflow from the output of katana and the contents of discovered URLs via curl (referenced in SKILL.md).
  • Boundary markers: None are defined in the command examples, as the skill typically passes data through UNIX pipes.
  • Capability inventory: The skill uses katana (crawler), nuclei (scanner), subfinder (recon), httpx (prober), and curl (network client).
  • Sanitization: The examples rely on specific string matching (e.g., grep) rather than general LLM processing, which limits the risk of the agent executing embedded instructions from the crawled content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 04:24 AM