100xAI Workflow
writing good code
Section titled “writing good code”coding agents
Section titled “coding agents”| Tool | What it is | Cost |
|---|---|---|
| Claude Code | Anthropic’s coding agent for the terminal, IDE, and web | Paid (Claude plan or API) |
| Codex | OpenAI’s coding agent for the terminal, IDE, and cloud | Paid (ChatGPT plan or API) |
| Cursor | AI code editor with agent mode and cloud agents | Free tier |
| Google Antigravity | Google’s agent-first IDE | Free tier |
| GitHub Copilot | Completions, agent mode in VS Code, and a cloud coding agent | Free tier |
| Devin | Cognition’s cloud agent that works in its own sandbox and opens pull requests; Windsurf, its editor, is now Devin Desktop | Paid |
| Amp | Sourcegraph’s coding agent for the terminal and editor | Paid |
| Factory | Droids, coding agents for the terminal, IDE, CI, and cloud | Free tier |
| Kiro | AWS’s spec-driven AI IDE and CLI | Free tier |
| Zed | Fast editor with built-in agent support | Free tier |
| Warp | Terminal with built-in coding agents and cloud agents | Free tier |
| Augment Code | IDE and CLI agent that indexes large codebases for context | Paid |
| JetBrains Junie | Coding agent inside JetBrains IDEs | Paid |
| Tabnine | Completions and agents that can run on your own servers | Paid |
| Google Jules | Async agent that works on a GitHub repo in a VM and opens pull requests | Free tier |
| Gemini CLI | Google’s terminal agent for Gemini models | Open source |
| Qwen Code | Terminal agent built for Qwen coder models | Open source |
| Kimi Code CLI | Moonshot AI’s terminal agent for Kimi models | Open source |
| ZCode | Z.ai’s coding agent for GLM models | Open source |
| Deep Code | Community terminal agent tuned for DeepSeek models; DeepSeek’s API docs cover setup | Open source |
| Pi | Minimal, extensible terminal coding agent that works with many model providers | Open source |
| Mistral Vibe | Mistral’s minimal CLI coding agent | Open source |
| OpenCode | Terminal coding agent, works with any model provider | Open source |
| Crush | Charm’s terminal coding agent, works with many model providers | Open source |
| Goose | Block’s local agent, works with any model provider | Open source |
| Aider | Terminal pair programmer that commits to git | Open source |
| Cline | Coding agent as a VS Code extension | Open source |
| Kilo Code | Coding agent for VS Code, JetBrains, and the CLI | Open source |
| Continue | IDE assistant and CLI that can also run agents as pull request checks | Open source |
running agents in parallel
Section titled “running agents in parallel”| Tool | What it is | Cost |
|---|---|---|
| Conductor | Mac app that runs several Claude Code and Codex agents at once, each in its own git worktree | Free app, agents billed separately |
| Capy | Cloud IDE that runs many coding agents on separate tasks, from task to merged pull request | Paid |
| bb | Workbench for running and coordinating Claude Code, Codex, Cursor, and other agents at once | Open source |
| Vibe Kanban | Kanban board that sends each task to a coding agent in its own worktree | Open source |
| Sculptor | Imbue’s desktop app that runs parallel coding agents in containers | Free |
| Nimbalyst | Desktop app, formerly Crystal, that runs Claude Code and Codex sessions side by side in worktrees | Free |
| Emdash | Desktop app that runs agents from different providers in parallel worktrees | Open source |
| Superset | Desktop app for running and switching between many CLI agents | Free tier |
| Claude Squad | tmux manager for several Claude Code, Codex, or Aider sessions in worktrees | Open source |
app builders
Section titled “app builders”| Tool | What it is | Cost |
|---|---|---|
| v0 | Vercel’s tool for generating React and Next.js apps and deploying them | Free tier |
| Bolt.new | Builds and runs full-stack web apps from prompts in the browser | Free tier |
| Lovable | Builds web apps from prompts and syncs the code to GitHub | Free tier |
| Replit Agent | Builds and deploys apps inside Replit’s hosted workspace | Free tier |
libraries and add-ons for your agents
Section titled “libraries and add-ons for your agents”| Tool | What it is | Cost |
|---|---|---|
| Context7 | MCP server that gives agents current library documentation | Free tier |
| Superpowers | Skills library that gives coding agents a workflow: brainstorm, plan, test first, review | Open source |
| Anthropic Skills | Example and reference agent skills from Anthropic | Open source |
| AGENTS.md | Shared file format for giving any coding agent project instructions | Free |
| Task Master | Breaks a spec into tasks that a coding agent works through | Open source |
| BMAD Method | Set of agent roles and workflows for planning and building with coding agents | Open source |
| Serena | MCP server that gives agents symbol-level code search and editing through language servers | Open source |
| Playwright MCP | Lets an agent drive a real browser to test what it built | Open source |
| Chrome DevTools MCP | Lets an agent inspect a live Chrome page, console, network, and performance | Open source |
| GitHub MCP Server | Gives agents access to issues, pull requests, and Actions | Open source |
| Repomix | Packs a repo into one file to paste into a model | Open source |
| Agent Reach | Lets an agent read and search Twitter, Reddit, YouTube, GitHub, and similar sites without API keys | Open source |
| Hermes Agent | Nous Research’s general agent with memory, skills, and scheduled jobs; usable for coding and automation | Open source |
reviewing your code
Section titled “reviewing your code”pull request review
Section titled “pull request review”| Tool | What it is | Cost |
|---|---|---|
| CodeRabbit | Line-by-line review on pull requests, plus a CLI and IDE extension | Free tier, free for open source |
| Cursor Bugbot | Pull request reviewer focused on finding bugs | Paid |
| GitHub Copilot code review | Request Copilot as a reviewer on a pull request | Included in Copilot plans |
| Claude Code GitHub Actions | Runs Claude on pull requests and issues from a workflow | Paid (API usage) |
| Codex code review | Codex reviews pull requests on GitHub | Paid (ChatGPT plan) |
| Graphite Agent | Reviewer that comments on pull requests, with stacked pull requests; Graphite is now part of Cursor | Free tier |
| Greptile | Reviewer that indexes the whole codebase for context | Paid |
| Open Code Review | Alibaba’s self-hosted reviewer that combines fixed checks with an LLM and leaves line comments | Open source |
| Qodo | Code review and test generation | Free tier |
| Ellipsis | Reviews pull requests and can push fixes to them | Paid |
| Sourcery | Reviews pull requests on GitHub and GitLab, and in the IDE | Free tier |
| cubic | Reviews pull requests and learns from team feedback | Free tier |
| Baz | Checks pull requests against the plan and the codebase; also has a local CLI | Paid |
| CodeAnt AI | Pull request review combined with static analysis and security scanning | Paid |
security and tests
Section titled “security and tests”| Tool | What it is | Cost |
|---|---|---|
| Snyk | Finds vulnerabilities in code and dependencies and suggests fixes | Free tier |
| Semgrep Assistant | Triages Semgrep findings, marks false positives, and suggests fixes | Free tier |
| Aikido | Scans code and dependencies and suggests fixes | Free tier |
| Corgea | Finds vulnerabilities and writes fixes for them | Paid |
| Codex Security | Scans a repo for vulnerabilities and proposes patches | Paid |
| Diffblue Cover | Writes unit tests for Java and Python | Paid |
| Tusk | Writes unit and integration tests for the code changed in a pull request | Paid |
| Tool | What it is | Cost |
|---|---|---|
| Blacksmith | Faster GitHub Actions runners; its Codesmith agent fixes failed checks | Free tier |
| GitHub Agentic Workflows | Runs coding agents as GitHub Actions from markdown workflow files | Open source, model usage paid |
| Nx Self-Healing CI | Analyzes failed CI tasks and proposes a fix for the pull request | Free tier |
| CircleCI Chunk | Fixes flaky tests and failing builds in CircleCI | Paid |
| Gitar | Analyzes CI failures and pushes fixes | Paid |
| Trunk | Flaky test detection, quarantine, and merge queue with AI failure summaries | Paid |
| Buildkite MCP server | Gives coding agents the logs of failed builds | Open source, Buildkite paid |
watching prod
Section titled “watching prod”errors, analytics, and observability
Section titled “errors, analytics, and observability”| Tool | What it is | Cost |
|---|---|---|
| PostHog | PostHog AI answers questions about analytics, errors, and replays; the MCP server gives your agent the same data; PostHog Desktop turns errors and tickets into pull requests | Free tier |
| Polylane | Agents that watch metrics, logs, and traces, open fix pull requests, and give production context to your coding agent | Free tier |
| Sentry Seer | Root cause analysis and suggested fixes for Sentry errors | Paid add-on |
| Grafana Assistant | Builds queries and dashboards and runs root cause investigations in Grafana Cloud | Free tier |
| Honeycomb | Canvas and an MCP server for querying traces while debugging | Free tier |
| Datadog Bits AI SRE | Investigates Datadog alerts and proposes causes | Paid |
| New Relic SRE Agent | Investigates alerts and incidents over New Relic data | Paid |
| Dynatrace | SRE, developer, and security agents over Dynatrace data | Paid |
| Coralogix Olly | Answers questions about logs, metrics, and traces in Coralogix | Paid |
| Dash0 Agent0 | Troubleshooting and query agents in an OpenTelemetry-native tool | Paid |
| Metoro | Kubernetes observability with an agent that investigates alerts and opens fix pull requests | Free tier |
| Amplitude AI Agents | Watch product analytics and suggest experiments and fixes | Paid |
incidents and on-call
Section titled “incidents and on-call”| Tool | What it is | Cost |
|---|---|---|
| incident.io AI SRE | Investigates incidents and drafts updates | Paid |
| PagerDuty AI agents | Triage and incident agents inside PagerDuty | Paid |
| Rootly AI SRE | Investigates in Slack and drafts timelines and postmortems | Paid |
| Better Stack AI SRE | Correlates logs, metrics, traces, and deploys, and opens fix pull requests | Free tier |
| Resolve AI | On-call agent across logs, metrics, and code | Paid |
| Cleric | Checks deploys, triages alerts, and hands off fix pull requests | Paid |
| Traversal | Diagnoses incidents and triages alerts in large systems | Paid |
| Causely | Builds a causal model of services to find root causes | Paid |
| NeuBird Hawkeye | Investigates incidents across cloud and observability tools | Paid |
| Deductive AI | Incident investigation over code and telemetry, now part of Elastic | Paid |
| AWS DevOps Agent | Investigates incidents across AWS resources | Paid |
| Azure SRE Agent | Diagnoses and fixes problems in Azure resources | Paid |
Kubernetes
Section titled “Kubernetes”| Tool | What it is | Cost |
|---|---|---|
| HolmesGPT | CNCF sandbox agent that investigates alerts with read-only Kubernetes and Prometheus access | Open source |
| K8sGPT | Scans a cluster and explains problems in plain language | Open source |
| Komodor Klaudia | Kubernetes troubleshooting agent | Paid |
| Parity | On-call assistant that runs root cause analysis and runbooks on Kubernetes alerts | Paid |
| Aurora | Investigates incidents across clouds and writes postmortems | Open source |
creating issues
Section titled “creating issues”trackers and planning
Section titled “trackers and planning”| Tool | What it is | Cost |
|---|---|---|
| Linear | Issue tracker with AI triage and agents you can assign issues to | Free tier |
| Jira with Rovo | Atlassian’s AI for writing and triaging Jira work items | Free tier |
| Plane | Tracker with agents that triage items and write specs from short notes | Open source, agents paid |
| Korey | Turns notes and Slack threads into stories with acceptance criteria | Free tier |
| Jam | Records a bug with console and network logs and files it to Jira or Linear | Free tier |
| Traycer | Writes a spec and plan from your codebase and hands tasks to coding agents | Free tier |
| GitHub Spec Kit | Templates and a CLI for spec-driven development with any agent | Open source |
issue to pull request
Section titled “issue to pull request”| Tool | What it is | Cost |
|---|---|---|
| GitHub Copilot coding agent | Assign a GitHub issue to Copilot and it opens a pull request | Paid Copilot plans |
| Codex cloud | Assign from a Linear issue or Slack message and get a pull request | Paid (ChatGPT plan) |
| Cursor cloud agents | Start from Slack, Linear, or GitHub and get a pull request | Paid |
| Claude Code in Slack | Starts a Claude Code session from a Slack thread and returns a pull request | Paid |
| GitLab Duo Agent Platform | GitLab agents, including an issue to merge request flow | Paid |
| ClickUp Codegen | Turns ClickUp tasks into pull requests | Paid |
| Tembo | Runs coding agents in cloud environments, started from issues or chat | Paid |
| Devin | Takes tickets from Slack, Linear, or Jira | Paid |
| Google Jules | Works from GitHub tasks | Free tier |
browsers and automation
Section titled “browsers and automation”agentic browsers
Section titled “agentic browsers”| Tool | What it is | Cost |
|---|---|---|
| Perplexity Comet | Browser with an assistant that reads pages and runs multi-step tasks in your tabs | Free tier |
| Claude in Chrome | Chrome extension that lets Claude navigate, click, and fill forms in your browser | Paid (Claude plan) |
| Gemini in Chrome | Gemini side panel in Chrome; auto browse runs multi-step web tasks | Free tier, auto browse paid |
| ChatGPT agent | Browses and does tasks in a remote browser, or in your own through the Chrome extension; replaced Atlas and Operator | Paid |
| Microsoft Edge | Copilot Actions act on pages and Journeys group past browsing, built into Edge | Free |
| Dia | macOS browser from The Browser Company with an assistant across tabs and reusable skills | Free tier |
| Opera Neon | Opera’s browser that runs tasks and builds small outputs from web content | Paid |
| Genspark AI Browser | Browser with an autopilot mode for web tasks and an MCP store | Free tier |
| Fellou | Browser that plans and runs multi-step web tasks and research across sites | Free tier |
| Sigma AI Browser | Privacy-focused browser with a built-in agent and optional local models | Free tier |
| Brave Leo | Brave’s assistant, with an early agent mode that acts on pages | Free tier |
| Manus | General agent that runs research, browsing, and file tasks in a cloud sandbox | Free tier |
automation
Section titled “automation”| Tool | What it is | Cost |
|---|---|---|
| n8n | Node-based workflow automation with AI agent nodes and self-hosting | Free to self-host, cloud paid |
| Zapier Agents | Agents that act across Zapier’s app integrations | Free tier |
| Make | Visual scenario builder with agents that call Make’s app modules | Free tier |
| Activepieces | Automation builder with AI agent steps and MCP servers | Open source, cloud paid |
| Pipedream | Code-first integration platform with app connectors and MCP servers for agents | Free tier |
| Windmill | Turns scripts into workflows, webhooks, and internal UIs, with AI agent steps | Open source |
| Trigger.dev | Durable background jobs and AI agents in TypeScript | Open source, cloud free tier |
| Temporal | Durable execution engine that keeps long agent runs going through failures | Open source, cloud paid |
| Dify | Builder for agent workflows, RAG pipelines, and LLM apps | Open source, cloud free tier |
| Langflow | Visual builder for agent and RAG flows in Python | Open source |
| Flowise | Visual builder for LLM agents and chat flows | Open source |
| Skyvern | Automates browser workflows such as forms and portals with LLMs and vision | Open source, cloud paid |
| Gumloop | No-code canvas for AI workflows that connect to business apps | Free tier |
| Lindy | AI assistants for email, meetings, CRM updates, and similar tasks | Free tier |
| Relevance AI | Teams of AI agents for sales and operations tasks | Free tier |
| Bardeen | Automates browser and CRM workflows for sales and marketing | Free tier |
| Claude Cowork | Claude’s desktop agent that works on local files and apps, now part of the Claude app | Paid |
| Google Opal | Chains prompts, models, and tools into small AI apps | Free |
| Google Workspace Studio | Automated flows across Gmail, Drive, Docs, and Chat | Paid (Workspace plans) |
| Microsoft Copilot Studio | Low-code builder for Microsoft 365 agents, including computer use | Paid |
| UiPath Agentic Automation | Enterprise RPA that coordinates robots, agents, and people | Paid |
building agents
Section titled “building agents”frameworks and SDKs
Section titled “frameworks and SDKs”| Tool | What it is | Language | Cost |
|---|---|---|---|
| Flue | Astro team’s framework for headless agents that run in a sandbox; deploys to Cloudflare | TypeScript | Open source |
| Claude Agent SDK | The loop, tools, context handling, and subagents behind Claude Code as a library | Python, TypeScript | Open source, API usage paid |
| OpenAI Agents SDK | Agents with tools, handoffs, guardrails, tracing, and sandboxed execution | Python, TypeScript | Open source |
| Codex SDK | Drives the Codex coding agent from your own code | TypeScript | Open source, API usage paid |
| Google ADK | Google’s Agent Development Kit for single and multi-agent systems | Python, Java, Go, TypeScript | Open source |
| LangGraph | Agents as stateful graphs with checkpoints and human-in-the-loop steps | Python, TypeScript | Open source, paid platform |
| LangChain | Model, tool, and retriever integrations with an agent API on LangGraph | Python, TypeScript | Open source |
| Microsoft Agent Framework | Agents and graph workflows; replaces AutoGen and Semantic Kernel | Python, .NET | Open source |
| CrewAI | Groups of agents with roles, plus event-driven flows | Python | Open source, paid platform |
| Pydantic AI | Type-safe agents with validated structured output | Python | Open source |
| Mastra | Agents, workflows, memory, RAG, and evals in one framework | TypeScript | Open source, paid cloud |
| Vercel AI SDK | Provider-neutral API for text, tool calls, streaming, and agent loops, with UI hooks | TypeScript | Open source |
| Strands Agents | AWS’s model-driven agent loop with tools and MCP | Python, TypeScript | Open source |
| LlamaIndex | Data connectors, indexing, and retrieval, with agent workflows over documents | Python, TypeScript | Open source, paid cloud |
| Agno | Agents and teams with memory, knowledge, and tools; formerly Phidata | Python | Open source, paid platform |
| smolagents | Small library for agents that write and run Python as their actions | Python | Open source |
| DSPy | Writes LM programs as modules and optimizes their prompts | Python | Open source |
| OpenHands SDK | Library for building coding agents, the engine behind OpenHands | Python | Open source, paid cloud |
| Cloudflare Agents SDK | Stateful agents on Durable Objects with scheduling, WebSockets, and MCP | TypeScript | Open source, usage billed |
| Inngest AgentKit | Multi-agent networks with routing and MCP tools on durable execution | TypeScript | Open source |
| VoltAgent | Agents, workflows, memory, and an observability console | TypeScript | Open source, paid console |
| Genkit | Google’s flows, tool calling, and local dev UI for AI features | TypeScript, Go, Python | Open source |
| BeeAI Framework | IBM’s agents and workflows, now under the Linux Foundation | Python, TypeScript | Open source |
| Haystack | Pipelines and agents for RAG and tool use | Python | Open source, paid platform |
| Rig | Provider-neutral LLM clients, embeddings, tools, and agents | Rust | Open source |
| Eino | ByteDance’s LLM components, graphs, and ReAct agent | Go | Open source |
protocols and tool layers
Section titled “protocols and tool layers”| Tool | What it is | Language | Cost |
|---|---|---|---|
| MCP SDKs | Official libraries for building MCP servers and clients | Most major languages | Open source |
| FastMCP | Decorator-based framework for MCP servers and clients | Python | Open source |
| A2A | Protocol and SDKs for agents to find and call each other | Python, TypeScript, Java, Go, .NET | Open source |
| AG-UI | Event protocol that streams agent state and actions to a frontend | TypeScript, Python | Open source |
| CopilotKit | React components and runtime for agent UIs inside your app | TypeScript | Open source, paid cloud |
| Composio | Hosted connectors to many SaaS APIs with managed auth | Python, TypeScript | Free tier |
| Arcade | Tool runtime with per-user OAuth for building and hosting MCP tools; now owns Smithery | Python, TypeScript | Free tier |
| Smithery | Registry and hosting for MCP servers | TypeScript | Free tier |
evals and tracing
Section titled “evals and tracing”| Tool | What it is | Language | Cost |
|---|---|---|---|
| Langfuse | Traces, prompt management, evals, and datasets; self-hostable | TypeScript, Python | Open source, paid cloud |
| LangSmith | Tracing, datasets, and evals for any framework | Python, TypeScript | Free tier |
| Arize Phoenix | Tracing and evals on OpenTelemetry; self-hostable | Python, TypeScript | Open source |
| Braintrust | Eval experiments, scorers, logging, and a prompt playground | Python, TypeScript | Free tier |
| Promptfoo | CLI for prompt and agent evals and red-teaming; now owned by OpenAI | TypeScript | Open source |
| Opik | Comet’s tracing, evals, and prompt optimization; self-hostable | Python, TypeScript | Open source, paid cloud |
| W&B Weave | Traces and scores LLM calls and agent runs | Python, TypeScript | Free tier |
| MLflow | Tracing and evals for LLM apps inside MLflow | Python | Open source |
memory, sandboxes, and browsers
Section titled “memory, sandboxes, and browsers”| Tool | What it is | Language | Cost |
|---|---|---|---|
| Mem0 | Extracts and retrieves long-term user and agent memories | Python, TypeScript | Open source, paid platform |
| Graphiti | Temporal knowledge graph for agent memory; Zep is the hosted version | Python | Open source |
| Letta | Stateful agents whose memory lives on a server; formerly MemGPT | Python, TypeScript | Open source, paid cloud |
| E2B | MicroVM sandboxes where agents run the code they write | Python, TypeScript | Open source, paid cloud |
| Daytona | Fast-starting sandboxes for agent code | Python, TypeScript | Open source, paid cloud |
| Modal | Serverless containers, GPUs, and sandboxes controlled from code | Python, TypeScript, Go | Paid, monthly free credits |
| Vercel Sandbox | Isolated microVMs for running agent code | TypeScript, Python | Free tier |
| Cloudflare Sandbox SDK | Runs code in Cloudflare Containers from a Worker | TypeScript | Open source, usage billed |
| Browserbase | Hosted headless browsers for agents | Python, TypeScript | Free tier |
| Stagehand | Browser automation that mixes code with natural-language steps | TypeScript | Open source |
| browser-use | Lets a model drive a real browser to finish web tasks | Python | Open source, paid cloud |