dev-agent

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill is configured to execute arbitrary commands defined in the project's YAML configuration under testing.commands. This typically includes development tasks like running tests (npm test), linting, or type-checking.
  • [PROMPT_INJECTION]: The agent is susceptible to indirect prompt injection because it ingests untrusted data from external sources such as Jira ticket descriptions and Slack/Teams messages. This content could potentially contain instructions aimed at influencing the agent's code generation or decision-making process.
  • Ingestion points: Task management tools (Jira, ClickUp, GitHub Issues) and communication platforms (Slack, Teams).
  • Capability inventory: File system writes (code implementation), subprocess execution (tests/linting), and network operations (creating Pull Requests, sending notifications).
  • Boundary markers: The skill includes an assessment phase to decide autonomy levels, which acts as a logical check before execution.
  • Sanitization: There is no explicit sanitization of ticket descriptions before they are processed by the planning or implementation agents.
  • [EXTERNAL_DOWNLOADS]: The skill performs repository cloning (git clone) using URLs provided in the local or shared project configuration files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 08:12 PM