sprint-executor

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill performs local syntax and build checks using standard tools like python, npx tsc, cargo check, and dotnet build. These operations are limited to verifying code quality within the local repository and are appropriate for the skill's technical lead role.
  • [PROMPT_INJECTION]: The skill processes external data from ticket trackers (JIRA, Linear, Azure DevOps) and project documentation. It mitigates potential indirect prompt injection by using structured templates for subagent delegation and maintaining a human-in-the-loop review process combined with automated quality checks.
  • [SAFE]: The skill implements a proactive security measure by invoking a vulnerability-manager whenever a diff modifies dependency manifests or lockfiles. This ensures that third-party library updates are checked for known vulnerabilities before being committed to the repository.
  • [EXTERNAL_DOWNLOADS]: The skill interacts with external issue tracking platforms (JIRA, Linear, Azure DevOps) using specific Model Context Protocol (MCP) tools. These connections are used exclusively for managing ticket lifecycles and posting implementation logs as intended by the skill's design.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 08:27 PM
Security Audit — agent-trust-hub — sprint-executor