company-intel

Pass

Audited by Gen Agent Trust Hub on Nov 10, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads company-wise problem datasets from a public GitHub repository (liquidslr/leetcode-company-wise-problems). This data is used to populate interview preparation briefs.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, including community-maintained CSV files and results from web searches (LeetCode Discuss, Glassdoor, Reddit). This represents an attack surface where instructions embedded in those sources could influence the agent's behavior.
  • Ingestion points: scripts/questions.py (via GitHub raw URLs) and web search tools.
  • Boundary markers: Lacks explicit delimiters or instructions to disregard embedded commands in the fetched data.
  • Capability inventory: File writes to $CRACKED_DIR, command execution of bundled scripts, and network access.
  • Sanitization: No evidence of content sanitization or filtering for natural language instructions in the downloaded data.
  • [COMMAND_EXECUTION]: The skill executes a bundled Python script (questions.py) to process data and uses shell commands (find) to locate its own file paths across various agent configuration directories.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 10, 2026, 06:30 AM