Development plugins

Project contracts and specialist procedures for coding agents. Includes 8 composable skills.

8 skills

Specialist subagents that take a noisy job and return one structured report: log-analyst (correlate logs to a root cause), codebase-archaeologist (map an unfamiliar or legacy system), performance-optimizer (measure, fix by impact, re-measure), and migration-specialist (breaking-change catalog and a phased plan with rollback points).

4 skills · 4 agents

Knowledge graph for token-efficient code reviews

6 skills · 1 MCP server · hooks

Structured development methodology commands supporting the Explore-Plan-Code-Commit cycle from initial discovery through final delivery

7 commands

Interactive Pallas TPU kernel optimization via an iterative think-submit-analyze-reflect loop on a Cloud TPU VM over SSH. Skills: init-kernel, profile-brief, start, submit, analyze, reflect, transfer.

7 skills

Official Plivo skills for coding agents.

7 skills

Pull-request review as a set of specialists: a read-only code explorer, a behaviour-preserving simplifier, hunters for silent failures and test gaps, a type-design reviewer, and a /review-pr command that runs them in parallel and merges one ranked review.

1 command · 5 agents

SchemeWeaver MCP server — Schema.org mapping + JSON-LD tools for AI assistants, bundling the schemeweaver-setup, schemeweaver-map and schemeweaver-audit skills

4 skills · 2 MCP servers

Operate SIMPO through SimpoCLI and create or reproduce reviewed BioModel, DataSet, and paper-to-Project workflows.

5 skills

Reduces token consumption by 60-90% via AST-aware lazy file reading, structural symbol navigation, and cross-session tool-usage analytics. 25 MCP tools + read gates that step aside when the tools are not in the session.

3 skills · 1 MCP server · hooks

Build a modern, themed (Light/Dark) WPF UI with Joufflu: skills for the API of every package, app shell scaffolding, custom themes and restyling existing views onto the design system.

5 skills

Local-first governance for AI coding agents: approved shared memory, bounded runtime evidence, Failure Lessons, project and provider isolation, architecture contracts, MCP, hooks, and release integrity.

5 skills

Missions inside Claude Code: give /mission a goal and Medley plans the work, then supervises execution until the goal is met — bringing you in for the higher-level decisions and reviews that actually need you, not routine approvals. Work is distributed across models: Claude by default, extending to GPT, open-source, and others you configure when that buys performance or token efficiency. Steer any time from Claude Code or a live dashboard, and everything stays on your machine — nothing leaves your local environment except model calls and optional debug info.

2 skills · 2 MCP servers · hooks

ออกแบบ PCB ด้วย KiCad: ตรวจ ERC/DRC ตามสเปกโรงงาน, export Gerber/BOM/CPL ให้ JLCPCB/PCBWay/โรงงานไทย, เช็กลิสต์ มอก./กสทช., หาชิ้นส่วนในไทย, ตัวหนังสือไทยบนซิลก์สกรีน

5 skills

Reviewed coding tasks with focused specialist agents, optional PR delivery, and durable autonomous project work.

5 skills

Multi-perspective code, architecture and security review agents (upstream: wshobson/agents, MIT; pinned, not copied)

2 commands · 3 agents

C4-model architecture documentation agents (context/container/component/code) (upstream: wshobson/agents, MIT; pinned, not copied)

1 command · 4 agents

ATC fixing in confirmed batches, clean-core audit, Clean ABAP review with cited rules, ABAP Unit runs, debugging with breakpoints on shared systems.

5 skills

Enrich Claude's knowledge for authoring high-quality GitHub contributions (PRs and Issues)

2 skills · 2 commands

Host-gated intent review and mutation testing.

3 skills · hooks

Route each subtask to the model that fits, then delegate through a compatible caller-owned native agent or vendor CLI.

1 skill · 2 commands · hooks

Advanced git operations -- rebasing, cherry-picking, conflict resolution, bisect, stash management, history exploration, and undo strategies. Agent: git-assistant. Hook: commit-lint.

1 skill · 1 command · 1 agent · hooks

Build, integrate, investigate Findings, search logs, and observe Kapso WhatsApp automations and Project Event workflows.

3 skills · 1 MCP server

Four focused subagents — researcher, planner, code-reviewer, debugger — for deep codebase research, dependency-aware planning, pre-commit review, and root-cause debugging. Adds no skill-listing context cost.

4 agents

Architecture, implementation, testing, and dependency policy.

2 skills · 1 command · 1 agent

Nuxt UI and Content, Drizzle ORM, and Playwright E2E skills

4 skills

Infographic prompts, image generation, and interactive animated visualizations

4 skills

Run, debug, screenshot, and interact with a Flutter app on an Android emulator. Gives the agent eyes (screenshots, accessibility tree) and hands (tap, hold, swipe, multi-touch pinch).

4 skills

Polish Flutter app UI screens for visual hierarchy, readability, and section clarity. First reads a representative reference screen (the home page or a polished page the user points at) to lock in the app's existing design tone, then drives the running app on the Android emulator via the `android-emulator` skill — screenshots, taps, swipes — to observe each target screen, makes targeted widget and `ThemeData` changes, and spawns an independent evaluator sub-agent to score the result on design quality, originality, craft, and functionality. Use this skill whenever the user wants to improve, polish, refine, critique, audit, or tighten the visual design of a Flutter screen — even when they don't say "polish" explicitly (e.g. "make this look better", "fix the layout on settings", "tighten up the home screen", "this screen feels off"). Stops and asks for a reference screen if the app's design tone can't be established with confidence.

4 skills

Preview a single Flutter widget in isolation using the built-in Flutter Widget Previewer (`flutter widget-preview`) — wires up a project-specific `previewWrapper` so the widget renders against the app's real theme and providers, launches the local preview server, and lets the agent take headless-Chrome screenshots of the live preview while a human follows along in their own browser. Use whenever the user wants to design, iterate on, dogfood, share, audit, or sanity-check a standalone Flutter widget — even when they don't say "previewer" (e.g. "let me see the new button", "show me dark mode for this card", "throw this widget on a preview", "iterate on the empty state", "preview just the badge", "what does this look like at tablet width"). For full-screen design polish on a running app, use `design-polish` + `android-emulator` instead. Skips: non-Flutter UIs, full-app QA flows, projects on Flutter < 3.35.

4 skills