implement

Pass

Audited by Gen Agent Trust Hub on Aug 3, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the Bash tool to perform essential project tasks, such as running build scripts (make megalinter-build) and updating dependency locks (uv lock). These actions are consistent with the skill's stated purpose as a developer implementation tool.
  • [EXTERNAL_DOWNLOADS]: The agent is instructed to use WebSearch and WebFetch to gather public metadata for new linters (rules, licenses, and configuration URLs). This external data is used solely to populate project descriptor files.
  • [DATA_EXFILTRATION]: No exfiltration patterns were detected. The skill's network capabilities are scoped to information gathering related to software dependencies and documentation.
  • [PROMPT_INJECTION]: The skill takes user input via $ARGUMENTS to define the implementation task. While this is an ingestion point for natural language instructions, the skill provides a structured process and specific project rules that the agent must follow, reducing the likelihood of successful adversarial influence.
  • [SAFE]: The skill includes explicit instructions to avoid modifying sensitive or generated files (e.g., Dockerfiles in flavors/, generated test files) and mandates the use of internal configuration and logging modules rather than general system calls.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 3, 2026, 06:55 PM
Security Audit — agent-trust-hub — implement