willmarple/harness-visualizer

willmarple/harness-visualizer

Agent Claude Code

Reviews implementation for architectural consistency, code quality, and project convention adherence. Read-only analysis with confidence-scored findings.

willmarple/harness-visualizer

Agent Claude Code

Orchestrates parallel implementation of specs using Agent Teams. Decomposes work into dependency-ordered waves, spawns and coordinates Sonnet teammates, enforces file ownership boundaries, and manages quality gates.

willmarple/harness-visualizer

Agent Claude Code

Implements production code per task assignment. Follows project conventions, respects file ownership boundaries, and runs quality checks after implementation.

willmarple/harness-visualizer

Agent Claude Code

Researches codebase patterns and external best practices. Uses local codebase search and Perplexity MCP for external lookups. Read-only.

willmarple/harness-visualizer

Agent Claude Code

Writes comprehensive tests for implemented code. Covers all code paths, edge cases, and error conditions using the project's test framework.

willmarple/harness-visualizer

Agent Claude Code

Analyzes Stryker mutation testing results for TypeScript projects and recommends test improvements.

willmarple/harness-visualizer

Agent Claude Code

Reviews code against architectural best practices and project conventions for Node.js + TypeScript projects (Vue 3 frontend, Express + Socket.IO backend, Vitest test suite). When a spec is in scope, also audits the implementation against the design.md and validates that every must-have requirement is exercised.

willmarple/harness-visualizer

Agent Claude Code

Tracks quality metrics over time across QA pipeline runs and identifies trends.

willmarple/harness-visualizer

Agent Claude Code

Generates comprehensive Vitest tests for Node.js + TypeScript projects (Vue 3 components, Express handlers, scanner modules, shared zod schemas) based on risk analysis. Emits requirement→test mappings for spec traceability when a spec is in scope.

willmarple/harness-visualizer

Agent Claude Code

Analyzes code changes to identify risks and prioritize testing efforts for Node.js + TypeScript projects (Vue 3 frontend, Express + Socket.IO backend). When a spec is in scope, anchors risks to specific requirements.

willmarple/harness-visualizer

Agent Claude Code

Authors ONE pattern or feature documentation file from a self-contained brief. Reads evidence files, fills the template with prose grounded in real code, vocab-lints, writes the doc. Never edits other files. Returns a one-paragraph status.

willmarple/harness-visualizer

Agent Claude Code

You are the Codebase Explorer on a research team for a specification generation pipeline. Your role is to explore the project codebase and document relevant patterns, existing implementations, and architectural conventions.

willmarple/harness-visualizer

Agent Claude Code

You are the Integration Researcher on a research team for a specification generation pipeline. Your role is to investigate API integrations, external dependencies, third-party constraints, and system boundaries.

willmarple/harness-visualizer

Agent Claude Code

You are the Requirements Analyst on a research team for a specification generation pipeline. Your role is to validate feasibility, identify edge cases, assess risks, and ensure requirements completeness.

willmarple/harness-visualizer

Agent Claude Code

You are the Research Lead for a specification generation pipeline. Your role is to orchestrate the research phase by distributing work to your teammates and synthesizing their findings into a coherent research synthesis.