AI agents and developer tools
Coding agents, MCP servers and developer tools that change how software gets built. Each card explains in plain words what the tool does, who it is for and what to watch out for.
100 references · Updated
Claude CodeAgents & devHand coding tasks to an agent in your terminal; extend it with skills, hooks, MCP and subagents.
Claude DesignAgents & devPrototype from your own design system, then export ZIP, PDF or PPTX, or hand off to Claude Code.
Google AI StudioAgents & devGenerate a full app UI with Gemini first, then wire auth and a database such as Supabase.
LovableAgents & devTurn a text brief into a working app or landing page without starting from code.
Playwright CLIAgents & devBrowser control from the command line with element snapshots, so an agent can check its UI cheaply.
StitchAgents & devExport a generated UI's DESIGN.md or hook an MCP so your coding agent builds from the real design.
UX PilotAgents & devFour design variations per prompt, from text, sketch or screenshot, then export to Figma or code.
VariantAgents & devDescribe an idea and scroll through endless design variations instead of judging one result.
AurelioAgents & devNet-worth tracking and planning with an AI advisor, all records kept on your own computer.
Elephant-GoldfishAgents & devA plugin that has an agent design and review software before any code gets written.
OpenDotsAgents & devAn open template for persistent agents with their own computers that ask approval in Slack.
PostizAgents & devOne calendar for scheduling posts across networks, built so agents can operate it too.
text-to-cadAgents & devDescribe a part in plain words and get a 3D CAD model, with a cad package on PyPI.
The Printing PressAgents & devGenerate an agent-first CLI from an app's API docs, with SQLite sync and offline search.
BuzzyAgents & devCall video and image models from Claude or Cursor through an MCP server and ready-made Skills.
RunableAgents & devPrompt one agent for a site, slides, app or video, and point it at a Figma design to follow.
10xAgents & devDescribe an iOS app in a prompt and get a native app back, with no code to start from.
Agency AgentsAgents & devSpecialist agent personas as prompt files with workflow and deliverables, for many tools.
Agent BeaconAgents & devCapture sessions from many coding agents and share approved learnings via MCP or skills.
Agent BrowserAgents & devA browsing tool that lets an agent open and operate web pages, including to test an app.
AgentiveAgents & devAI agent platform
AgentMailAgents & devGive an agent its own inbox via API or MCP, with threads, webhooks and custom domains.
AI Job SearchAgents & devA Claude Code workflow for the job hunt: shortlist, tailored CV and cover letter, pipeline.
AlfredAgents & devHotkey launcher: a custom URL search opens ChatGPT or Claude, and snippets expand saved prompts.
AntigravityAgents & devAn app that runs several local agents in parallel, plus CLI, IDE and Python SDK.
ApifyAgents & devPick a ready-made scraper (actor) from a store and get structured web data as JSON, via API or MCP.
ARTEMISAgents & devDrive a real Android phone from an agent via MCP, with Logcat and screenshots collected for you.
Autoresearch (Karpathy)Agents & devAn agent loop: edit one file, train for 5 minutes, keep the change only if the metric improves.
BeeftextAgents & devStore three to five go-to prompts as keyword snippets that expand anywhere on Windows.
Claude CoworkAgents & devA work folder with CLAUDE.md, skills, connectors and scheduled tasks for recurring document work.
Claude Fable 5.1Agents & devA model aimed at long, unsupervised coding runs and at finding the root cause of bugs.
CLI-AnythingAgents & devA pipeline that analyzes a program and generates an agent-ready CLI with tests, docs and a SKILL.md.
CursorAgents & devAgents take tasks in separate chats, with Bugbot review and lint and tests as feedback.
ElicitAgents & devA literature finder for research papers, a first step before organizing sources elsewhere.
FirecrawlAgents & devOne API call turns a web page into clean Markdown or JSON, with crawl, search and URL mapping.
GbrainAgents & devA self-hosted memory for your agent: Markdown as source, hybrid search, shared over MCP.
Gemini Fullstack LangGraph QuickstartAgents & devA complete starter wiring a Gemini and LangGraph research agent to a React front end.
Gemma 4Agents & devAn open family in five sizes, from phone to PC, built for reasoning and agentic workflows.
GitDiagramAgents & devTurn a code repository into a diagram of its structure.
GitReverseAgents & devPaste a repository or site and get a prompt that describes it; swap hub for reverse in a GitHub URL.
GLM-5.3 FlashAgents & devA model that takes video, image, text and files, with 1M-token context and JSON output.
Google Workspace CLIAgents & devOne CLI over Drive, Gmail, Calendar and Sheets that an agent can call, with --dry-run, JSON out.
GranolaAgents & devMeeting notes from your device's audio, no bot on the call, plus an MCP to query them from AI tools.
Hermes AgentAgents & devA persistent agent on your own server that writes its own skills and replies on Telegram or Slack.
Jev UltrafastAgents & devA browser agent that acts on a numbered table of page elements instead of screenshots.
Kimi K3 in CAgents & devA readable C99 reference for running a huge MoE model on CPU with MXFP4 quantization.
LightRAGAgents & devRetrieval over a knowledge graph plus vectors, so answers can follow relationships between entities.
LinearAgents & devAn MCP server so an agent can read and update issues, projects and docs after OAuth sign-in.
LM StudioAgents & devRun local models, then point Claude Code at its Anthropic-compatible /v1/messages server.
MakeAgents & devBuild automations by prompt or drag and drop across 3,500+ apps, and run them from an AI assistant.
ManusAgents & devDescribe a result and an agent builds slides, sites, games or research, refined in conversation.
MLXAgents & devRun and train models on Apple silicon Macs with a NumPy-style Python API.
Muse CodeAgents & devA terminal and CI coding agent with sandbox and approvals from run one, plus a headless mode.
n8nAgents & devChain a trigger, an AI Agent node, a chat model, memory and tools into a workflow on a canvas.
NetlifyAgents & devPush to Git for a preview URL on every pull request, with one-click rollback and automatic HTTPS.
NotebookLMAgents & devLoad your research sources into one notebook and ask questions across all of them.
ObsidianAgents & devPlain Markdown files with links between notes: a folder an AI agent can read and write as a wiki.
Obsidian Local REST APIAgents & devExpose your Obsidian vault via a local REST API and MCP server so scripts and AI clients edit notes.
OllamaAgents & devRun open language models on your own machine and serve them locally to agents.
OmniRouteAgents & devOne local OpenAI-compatible endpoint that routes your coding tools across providers, with fallback.
OpalAgents & devDescribe a workflow in plain language and get an AI mini app whose prompts and models you can edit.
Open NotebookAgents & devA self-hosted research notebook: chat with your sources using any model, local or API.
Open SEOAgents & devSelf-host keyword research, rank tracking, backlinks and site audits, with an MCP server for agents.
OpenClawAgents & devAn assistant on your own machine you message from WhatsApp or Slack, with roles in SOUL.md files.
OpenCodeAgents & devSeparate agents for building (full access) and planning (read-only), switched with Tab.
OpenRouterAgents & devOne API key and one credential to swap the model behind an n8n agent node.
OpenVikingAgents & devAgent memory and skills in a virtual file system you can ls, grep and edit, with tiered summaries.
OrcaAgents & devSend one prompt to several coding agents, each in its own git worktree, then compare and merge.
PaperclipAgents & devRun agents like a company: roles, goals, per-agent budgets and an inbox where you approve hires.
PerplexityAgents & devResearch a topic and get a finished piece back: dashboards, pages or quizzes built from the answer.
Personal AI Infrastructure (PAI)Agents & devAn architecture for a personal agent: memory, hooks, security and autonomy laid out in PAI/TELOS.
Personal Context PortfolioAgents & devTen markdown files about who you are and how you work, so any AI agent starts with your context.
PineconeAgents & devA managed vector index with metadata filters and namespaces, so an agent can recall by meaning.
Pizza BotAgents & devAn inbox for background agents: finished work in one queue, approvals in another.
PlaywrightAgents & devOne API to drive Chromium, Firefox and WebKit, with an MCP server so agents can browse and test.
Poppy AIAgents & devA chat panel that writes scripts and newsletters in your voice from boards of saved context.
PostmanAgents & devAPI skills and a CLI that hand a coding agent structured context about your APIs.
PowerToysAgents & devWindows launcher and window tools: Run, Command Palette, FancyZones, Color Picker, OCR text grab.
PromptmetheusAgents & devA prompt workbench for iterating on versions and comparing models side by side in A/B tests.
QMDAgents & devLocal search over your Markdown notes: keyword, semantic and re-ranked, from CLI or MCP.
Qwen 3.5Agents & devOpen models you can run on your own machine through Ollama or LM Studio, even behind Claude Code.
Qwen3.8-MaxAgents & devA repo describing a 2.4T-parameter MoE model with 1M-token context and listed per-token prices.
RaycastAgents & devKeyword snippets that paste your prompts, and quicklinks that open searches in ChatGPT or Claude.
REAAgents & devReverse-engineering tools (Ghidra, IDA, Hopper) as an MCP server for binaries, APKs and Electron.
Relay (Obsidian plugin)Agents & devObsidian plugin for sharing a vault's context with agents
Relevance AIAgents & devPrebuilt sales agents like Outbound Prospector and Deal Reviewer, with model choice and audit logs.
StrixAgents & devAn agent that runs your app, tries real exploits and confirms findings with proof-of-concept code.
SupabaseAgents & devA Postgres project with auth, storage, edge functions and REST/GraphQL APIs, no server to set up.
TavilyAgents & devWeb search for an agent over MCP, returning sources; pair with Firecrawl to read pages as Markdown.
Ternary Bonsai 2Agents & devLanguage models with weights stored as -1, 0 or +1: 0.43 GB for 2B, small enough for a phone.
Trail CodeAgents & devA tree-sitter code map plus markdown learnings versioned with the repo, so agents stop re-exploring.
TranscriptAPIAgents & devYouTube transcripts, search and channel lists from a REST API or MCP server, no YouTube quota.
Unsloth StudioAgents & devA desktop app to chat with, fine-tune and export local models, with an OpenAI-compatible API.
VercelAgents & devPush to a preview URL for every change, with an AI Gateway and sandbox for agent-built apps.
Vibe ProspectingAgents & devEnrich a company spreadsheet with decision-makers' names, titles, LinkedIn profiles and contacts.
ViewMaxAgents & devShort-video generation across model families, with a remote MCP server for Claude or Cursor.
VoiceflowAgents & devBuild chat and voice agents from playbooks, workflows and a knowledge base; test before launch.
Wispr FlowAgents & devDictate into any text field and get cleaned text back: filler words dropped, lists formatted.
Zapier MCPAgents & devOne MCP server that gives an assistant 9,000+ apps, with you choosing which actions it may use.
ZattenAgents & devA white-label WhatsApp setup of AI agent, CRM and kanban funnel to resell to an agency's clients.