Reads a repository as its current architecture, runs the capacity math, and compares structurally distinct candidates before recommending one. Outputs an interactive HTML report, a structured design.json, and a durable DESIGN.md.
Skills-first research workflows for Research Software Engineers and Researchers: research, plan, experiment, implement, validate, reproduce, harden, and handoff — interactive when you want collaboration, direct when you don't.
Agents and skills for Scientific Python development following community best practices. Covers packaging, testing, code quality, environment management with pixi, and documentation architecture.
Teaches the agent when and how to reach for the peekie CLI for Xcode .xcresult bundles. Ships as an umbrella skill plus one dedicated sub-skill per subcommand (peekie-tests, peekie-warnings, peekie-errors, peekie-coverage, peekie-attachments) — fully-qualified test names, attachment extraction, JSON output for jq pipelines, SonarQube reports
Graph-powered codebase reasoning, impact analysis, and governed edits via the GraQle MCP server, plus governed-workflow skills and an optional governance gate hook.
LSP and MCP support for DTCG design tokens - intelligent autocomplete, hover docs, diagnostics, code actions, and AI-native token discovery, validation, search, and conversion
Delegate to Gemini, Codex, Copilot, Cursor and opencode as sub-agents from Claude Code, on subscriptions you already pay for. Adds /agent-intern:second-opinion, :image and :doctor.
AI-assisted spec crafting through research, interviews, and multi-LLM review. Like Geppetto carved Pinocchio from wood, transform rough ideas into living implementation plans.
45 Claude Code skills and 7 specialist agents that run on your real code in any stack: ORM query optimization, API and architecture scaffolding, unit and integration tests, auth, caching, OpenTelemetry, logging, Docker and CI/CD, plus code review, architecture, security, database and observability audits.