agent-dev-coder

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its integration with web-searching and web-fetching tools.
  • Ingestion points: The agent ingests data from external sources using tools like mcp__brave__brave_web_search and WebFetch.
  • Boundary markers: The instructions lack explicit delimiters or instructions to ignore embedded commands within processed external data.
  • Capability inventory: The agent has broad permissions including Bash execution, file modification (Write, Edit, MultiEdit), and specialized coding tools like NotebookEdit.
  • Sanitization: There is no evidence of input validation or sanitization for data retrieved from the web before it is processed or used in terminal commands.
  • [COMMAND_EXECUTION]: The skill utilizes the Bash tool to perform build and test operations as part of its core mandate. While intended for legitimate development tasks, this provides a mechanism for arbitrary command execution if the agent is misled by malicious external input or poisoned project files.
  • [EXTERNAL_DOWNLOADS]: The skill incorporates WebFetch and Brave Search capabilities, allowing it to retrieve and process content from remote URLs and search results.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 01:13 AM