{
  "schema": "industrynext.ai-tools-mcp-atlas/v1",
  "id": "industrynext/toolchain-atlas-004",
  "title": "AI + MCP Toolchain Atlas",
  "canonical": "https://www.industrynext.xyz/ai-tools-mcp/",
  "published": "2026-08-31",
  "last_reviewed": "2026-08-31",
  "summary": "A practical, non-affiliate shortlist and local stack builder for selecting AI work surfaces, MCP bridges, permission posture, and verification evidence.",
  "editorial_policy": {
    "top_definition": "A compact shortlist selected for practical coverage and useful workflows, not a universal ranking.",
    "affiliate_compensation": false,
    "transmits_or_stores_builder_selections": false,
    "pricing_claims_included": false,
    "default_permission_posture": "Project-scoped configuration, least privilege, read-only first, and human approval before writes.",
    "warning": "Plans, availability, models, prices, data terms, transports, and server tools change. Confirm current terms and permissions with the official publisher before connecting accounts, spending money, or sharing sensitive material."
  },
  "definitions": {
    "ai_tool": "A work surface used to think, code, research, design, or make an artifact.",
    "mcp_server": "A permissioned bridge that can expose external context or actions to an MCP-capable host.",
    "receipt": "Inspectable evidence such as a source trail, diff, test, screenshot, log, or public URL."
  },
  "ai_tools": [
    {
      "id": "chatgpt",
      "name": "ChatGPT",
      "category": "think",
      "role": "General workbench",
      "use_when": "The work changes shape as you learn.",
      "receipt": "Sources, assumptions, and the decision.",
      "official_url": "https://chatgpt.com/"
    },
    {
      "id": "claude",
      "name": "Claude",
      "category": "think",
      "role": "Document workbench",
      "use_when": "The source packet is the work.",
      "receipt": "A claim-to-source table.",
      "official_url": "https://claude.ai/"
    },
    {
      "id": "gemini",
      "name": "Gemini",
      "category": "think",
      "role": "Google workbench",
      "use_when": "The approved workflow is already Google-shaped.",
      "receipt": "An export outside the chat.",
      "official_url": "https://gemini.google.com/"
    },
    {
      "id": "notebooklm",
      "name": "NotebookLM",
      "category": "think",
      "role": "Source-set workbench",
      "use_when": "The boundary is a known source set.",
      "receipt": "The source collection and cited notes.",
      "official_url": "https://notebook.google/"
    },
    {
      "id": "perplexity",
      "name": "Perplexity",
      "category": "think",
      "role": "Web scout",
      "use_when": "You need the current source landscape.",
      "receipt": "Opened primary links, not citation count.",
      "official_url": "https://www.perplexity.ai/"
    },
    {
      "id": "codex",
      "name": "Codex",
      "category": "code",
      "role": "Repository agent",
      "use_when": "The artifact lives in a real codebase.",
      "receipt": "A small diff and exact test output.",
      "official_url": "https://developers.openai.com/codex/"
    },
    {
      "id": "claude-code",
      "name": "Claude Code",
      "category": "code",
      "role": "Terminal agent",
      "use_when": "You want an agent in the shell loop.",
      "receipt": "The plan, diff, and verification transcript.",
      "official_url": "https://code.claude.com/docs/en/overview"
    },
    {
      "id": "cursor",
      "name": "Cursor",
      "category": "code",
      "role": "Editor agent",
      "use_when": "The editor is your control surface.",
      "receipt": "A reviewed commit, not pasted code.",
      "official_url": "https://cursor.com/"
    },
    {
      "id": "figma-make",
      "name": "Figma Make",
      "category": "make",
      "role": "Interface prototype",
      "use_when": "The unknown is interaction and hierarchy.",
      "receipt": "Prototype, observer notes, and handoff.",
      "official_url": "https://www.figma.com/make/"
    },
    {
      "id": "runway",
      "name": "Runway",
      "category": "make",
      "role": "Moving image",
      "use_when": "The deliverable needs motion.",
      "receipt": "A take sheet with keep and reject reasons.",
      "official_url": "https://runway.com/"
    }
  ],
  "mcp_servers": [
    {
      "id": "openai-docs",
      "name": "OpenAI Docs",
      "transport": "streamable HTTP",
      "access": "read-only public documentation",
      "use": "Search and read official OpenAI developer documentation in the agent context.",
      "first_test": "Find one current API field and return the exact official source page.",
      "boundary": "Documentation-only; it does not call the OpenAI API on the user's behalf.",
      "official_url": "https://learn.chatgpt.com/learn/docs-mcp",
      "codex_setup": "codex mcp add openaiDeveloperDocs --url https://developers.openai.com/mcp"
    },
    {
      "id": "context7",
      "name": "Context7",
      "transport": "local stdio process",
      "access": "documentation retrieval",
      "use": "Bring current library and framework documentation into a coding session.",
      "first_test": "Resolve one package and retrieve the current method signature.",
      "boundary": "Inspect the package and source before running it locally.",
      "official_url": "https://github.com/upstash/context7",
      "codex_setup": "codex mcp add context7 -- npx -y @upstash/context7-mcp"
    },
    {
      "id": "figma",
      "name": "Figma",
      "transport": "local or remote; see publisher guide",
      "access": "account and file scope; tools vary",
      "use": "Expose structured design context and supported Figma actions.",
      "first_test": "Read one named frame and summarize its components.",
      "boundary": "Start with one file and inspect every offered tool.",
      "official_url": "https://developers.figma.com/docs/figma-mcp-server/"
    },
    {
      "id": "playwright",
      "name": "Playwright",
      "transport": "local stdio process",
      "access": "browser control",
      "use": "Inspect and operate a browser to test interaction, layout, errors, and responsive behavior.",
      "first_test": "Open a local page, read the title, and take a screenshot.",
      "boundary": "Use a test profile; browser actions can change external state.",
      "official_url": "https://github.com/microsoft/playwright-mcp"
    },
    {
      "id": "chrome-devtools",
      "name": "Chrome DevTools",
      "transport": "local stdio process",
      "access": "browser and debugging control",
      "use": "Inspect a live browser's DOM, console, network, and performance state.",
      "first_test": "Read console errors from a local or staging route.",
      "boundary": "Avoid personal profiles and authenticated production tabs.",
      "official_url": "https://github.com/ChromeDevTools/chrome-devtools-mcp"
    },
    {
      "id": "sentry",
      "name": "Sentry",
      "transport": "see publisher guide",
      "access": "account and organization data",
      "use": "Bring issue and error context into a debugging session.",
      "first_test": "Read one test-project issue without changing its state.",
      "boundary": "Scope the organization and project; treat traces as sensitive.",
      "official_url": "https://docs.sentry.io/product/sentry-mcp/"
    },
    {
      "id": "github",
      "name": "GitHub",
      "transport": "local stdio or supported host setup",
      "access": "account; read and write toolsets",
      "use": "Work with repositories, issues, pull requests, actions, and security data.",
      "first_test": "Use read-only mode to summarize one issue and linked pull request.",
      "boundary": "Enable only needed toolsets; writes require explicit approval.",
      "official_url": "https://github.com/github/github-mcp-server"
    }
  ],
  "stack_builder": {
    "selection_location": "browser only",
    "share_encoding": "URL fragment containing only outcome and posture",
    "outcomes": [
      { "id": "site", "name": "Ship a useful website", "ai_tools": ["codex", "cursor"], "mcp_servers": ["github", "playwright", "sentry"] },
      { "id": "brief", "name": "Produce a technical brief", "ai_tools": ["chatgpt", "notebooklm"], "mcp_servers": ["openai-docs", "context7", "github"] },
      { "id": "design", "name": "Build from a design", "ai_tools": ["figma-make", "codex"], "mcp_servers": ["figma", "github", "playwright"] },
      { "id": "bug", "name": "Fix a production bug", "ai_tools": ["codex", "claude-code"], "mcp_servers": ["sentry", "github", "chrome-devtools"] },
      { "id": "review", "name": "Review a repository", "ai_tools": ["claude", "codex"], "mcp_servers": ["github", "context7", "playwright"] },
      { "id": "api", "name": "Prototype an AI API", "ai_tools": ["chatgpt", "codex"], "mcp_servers": ["openai-docs", "context7", "sentry"] }
    ],
    "postures": [
      { "id": "public", "name": "Public material and test data" },
      { "id": "team", "name": "Private team material" },
      { "id": "production", "name": "Production systems" }
    ]
  },
  "permission_gate": [
    "Name the receipt.",
    "Verify the publisher.",
    "Inventory every offered tool.",
    "Reduce the scope.",
    "Run a harmless read-only test.",
    "Keep a human approval and stop rule for consequential actions."
  ],
  "primary_sources": [
    "https://developers.openai.com/codex/mcp/",
    "https://learn.chatgpt.com/learn/docs-mcp",
    "https://modelcontextprotocol.io/",
    "https://registry.modelcontextprotocol.io/",
    "https://github.com/github/github-mcp-server"
  ]
}