QE-Agent
Agentic Development Lifecycle (ADLC) for Quality Engineering
execution
Jira → Context → Spec Gate → Implement → Code Gate → PR → CI → Validate → RCA → Loop- Built a multi-agent QE system with sub-agent orchestration that takes Jira tickets from context gathering and specification through implementation, review, CI validation, RCA, and iterative fixes.
- Orchestrated specialized agents, skills, and tools using Claude Code in headless mode, LiteLLM, Atlassian CLI, GitHub, Jenkins, Harness, Docker, CI/CD APIs, SSH, and custom Python/Shell automation across repositories, build pipelines, and live on-premise clusters.
- Added independent specification and code-review quality gates covering requirement understanding, coverage, reuse, correctness, framework fit, suite tagging, shared-utility safety, and feasibility.
- Added RCA and validation agents that distinguish product regressions from test failures, draft defect-ready findings, and feed validation failures back into the execution loop.
Multi-agent SystemsADLCClaude CodeLiteLLMPythonShellAtlassian CLIGitHubJenkinsHarnessDockerOpenSandboxCI/CD