Quality

Finishing A Development Branch helps you prepare completed work for merge with cleanup, testing, linting, documentation updates, and a structured pull request. Use it before pushing a branch for review.

Test-Driven Development uses a red-green-refactor cycle before implementation code is written. Use it for new features, bug fixes, and behavior-changing refactors.

Elf Inspection guides inspection of Linux ELF executables and shared libraries with tools like readelf, objdump, nm, and ldd. Use it to check symbols, sections, dependencies, debug info, build IDs, relocations, and linker or runtime symbol errors.

ML Debugging systematically diagnoses ML and AI training failures such as OOMs, NaNs, divergence, crashes, slow throughput, wrong outputs, and dependency conflicts. Use it when an experiment or model run breaks and you need root-cause analysis before applying fixes.

ML Verification checks code, configs, and math before training or deployment to catch mistakes early. It is used to validate framework behavior and flag issues before expensive runs.

Acceptance Criteria Review turns requirements and user stories into unambiguous, verifiable acceptance criteria. Use it to review plans or evidence for missing rules, failure paths, and testability gaps.

Accessibility Testing helps you design practical checks for WCAG, keyboard navigation, screen readers, and related assistive tech scenarios. Use it when planning or reviewing a11y tests for a real project.

Adversarial Review runs a stricter Claude Code review of local git changes and focuses on design, tradeoffs, and risky assumptions. Use it when you want stronger scrutiny than a normal review or a custom focus on a specific concern.

Agent Tool Testing verifies an AI agent’s tool selection, argument contracts, authorization, failure recovery, and side-effect boundaries. Use it when reviewing plans, results, or evidence for tool-using agents.

AI Agent Testing verifies agent planning, memory, tool use, recovery, safety boundaries, and task completion quality. Use it to review an agent run, assess evidence, and produce bounded improvements when context is incomplete.

AI-Assisted Testing helps with AI-driven QA workflows like test data generation, root-cause analysis, and prioritization. Use it when you need executable testing output and clear human verification points.

AI Feature Testing helps verify AI-enabled product features for quality, safety, fairness, fallback behavior, and user impact. Use it to review plans or results and produce bounded, evidence-based recommendations.