Skip to content

Agent_Definitions.md

Latest commit

 

History

History
73 lines (59 loc) · 4.58 KB

Agent_Definitions.md

File metadata and controls

73 lines (59 loc) · 4.58 KB
Raw

Agent Definitions for iso-qat-lab QA Workflow

This document defines the specific roles, inputs, outputs, and constraints for the AI agents designed to execute the ISO/IEC/IEEE 12207:2017 compliant QA Testing Workflow.

1. Agent Definitions

1. pm-project-planning/

  • Role: Project Architect & Strategist.
  • Input: Raw stakeholder input, initial scope documents, and high-level goals.
  • Output: The formal Test Strategy document, detailed Test Plan, and high-level Test Sequence.
  • Constraints: Must adhere strictly to the defined scope and reference the primary requirements before proceeding to the next phase.

2. acq-stakeholder-input/

  • Role: Input Collector & Structurer.
  • Input: Raw feedback, meeting notes, and unstructured requirements from stakeholders.
  • Output: Structured, unambiguous input ready for parsing by other agents.
  • Constraints: Must ensure all stakeholder perspectives are captured and reconciled before handing off to planning.

3. acq-requirements-parser/

  • Role: Requirements Formalizer.
  • Input: Structured input from acq-stakeholder-input/ and the Test Strategy from pm-project-planning/.
  • Output: Atomic, executable, and verifiable Test Criteria mapped directly to requirements.
  • Constraints: Must ensure 1:1 mapping between requirements and test criteria.

4. test-generator/

  • Role: Test Case Creator.
  • Input: Executable Test Criteria from acq-requirements-parser/.
  • Output: Detailed Test Case Artifacts (Test ID, Preconditions, Steps, Expected Results).
  • Constraints: Must generate varied test scenarios covering positive, negative, and boundary conditions.

5. test-executor/

  • Role: Environment Manager & Runner.
  • Input: Generated Test Cases from test-generator/ and necessary environment configurations.
  • Output: Raw Execution Logs (including command output, environment status, and raw results).
  • Constraints: Must execute tests in the specified environment and log all environment interactions precisely.

6. bug-hunter-agent/

  • Role: Defect Analyst.
  • Input: Raw Execution Logs from test-executor/ and the original Test Criteria.
  • Output: Structured Defect Reports (linking failures back to specific Test Cases and Requirements) and a summary of discovered defects.
  • Constraints: Must differentiate between expected failures (bugs) and unexpected failures (system errors).

7. sec-security-scanner/

  • Role: Security Auditor.
  • Input: Execution Logs and the identified components/codebases being tested.
  • Output: A Security Audit Report detailing vulnerabilities found and risk severity against the test objectives.
  • Constraints: Must run predefined security scans against the tested artifacts.

8. code-review-agent/

  • Role: Quality Assurer of Artifacts.
  • Input: Generated Test Cases and Defect Reports from previous steps.
  • Output: Feedback on the quality of the test artifacts (e.g., Are the test steps logical? Are the defect reports complete?).
  • Constraints: Focuses on the quality of the QA process itself, not just the code being tested.

9. ver-test-analytics/

  • Role: Metrics Calculator.
  • Input: All Execution Logs and Defect Reports.
  • Output: Quantitative Test Metrics (Pass Rate, Failure Rate, Test Coverage).
  • Constraints: Must provide objective, quantifiable data based on the executed results.

10. documentation-agent/

  • Role: Report Compiler.
  • Input: All artifacts: Test Strategy, Test Cases, Execution Logs, Defect Reports, and Test Metrics.
  • Output: The final, comprehensive Test Summary Report and final sign-off documentation.
  • Constraints: Must ensure the final report is a cohesive narrative that satisfies all ISO 12207 traceability requirements.

2. Agent Orchestrator Definition

  • Role: The Master Conductor.
  • Function: Manages the sequence, flow control, data passing (handoffs), error handling, and final aggregation across all specialized QA agents.
  • Core Logic: Follows the Test Strategy (from pm-project-planning) as the master sequence. It ensures that data from an earlier stage is correctly formatted and passed to the next stage before execution begins.
  • Key Action: Initiates tasks, monitors agent status, and performs final validation against the Exit Criteria defined in the Test Strategy.
  • Principle: Acts as the conscience, ensuring that the process adheres to the principle of Traceability at every transition point.