{
  "schema_version": "rangoon.article.v1",
  "id": "https://rangoon.ai/insights/archive-lovable-2-april-2025/",
  "url": "https://rangoon.ai/insights/archive-lovable-2-april-2025/",
  "markdown_url": "https://rangoon.ai/insights/archive-lovable-2-april-2025/index.md",
  "json_url": "https://rangoon.ai/insights/archive-lovable-2-april-2025/index.json",
  "title": "Lovable 2.0 added collaborative workspaces, a non-editing chat agent and a publish-time security scan",
  "summary": "Lovable announced Lovable 2.0 on April 24, 2025, combining workspaces, a planning and debugging Chat Mode Agent, Supabase-linked Security Scan and direct code editing.",
  "language": "en",
  "author": {
    "name": "Rangoon Editorial",
    "url": "https://rangoon.ai/insights/editorial/"
  },
  "publisher": {
    "name": "Rangoon",
    "maintainer": "Hypler",
    "url": "https://rangoon.ai/"
  },
  "format": "News analysis",
  "topic": {
    "slug": "agents",
    "label": "Agent systems",
    "url": "https://rangoon.ai/insights/topics/agents/"
  },
  "tags": [
    {
      "slug": "lovable",
      "label": "Lovable",
      "url": "https://rangoon.ai/insights/tags/lovable/"
    },
    {
      "slug": "lovable-2-0",
      "label": "Lovable 2.0",
      "url": "https://rangoon.ai/insights/tags/lovable-2-0/"
    },
    {
      "slug": "ai-software-collaboration",
      "label": "AI software collaboration",
      "url": "https://rangoon.ai/insights/tags/ai-software-collaboration/"
    }
  ],
  "dates": {
    "published": "2026-10-24",
    "updated": "2026-10-24",
    "event": "2025-04-24"
  },
  "event_source_id": 1,
  "retrospective": true,
  "reading_minutes": 3,
  "takeaways": [
    "Lovable announced Lovable 2.0 on April 24, 2025 with workspace collaboration, a Chat Mode Agent that planned and debugged without directly editing code, and a Supabase-linked Security Scan.",
    "The release separated roles and credit pools by plan, let teams invite project collaborators, and retained a Dev Mode for direct code editing alongside visual style changes.",
    "A publish-time scan and agentic planning could surface useful issues, but neither release demonstrated full security coverage, correct database policy, secure secrets or production readiness."
  ],
  "source_note": "Historical reporting: Source 0 is Lovable’s dated April 24, 2025 Lovable 2.0 release. Source 1 is Lovable’s dated February 12, 2025 Visual Edits release, used only as pre-release context for the direct-manipulation capability called out again in 2.0. It does not establish 2.0 availability or later platform terms.",
  "sources": [
    {
      "id": 1,
      "title": "Lovable Blog: Lovable 2.0: What’s New and What It Means for no-code App Builders",
      "url": "https://lovable.dev/blog/lovable-2-0",
      "published": "2025-04-24"
    },
    {
      "id": 2,
      "title": "Lovable Blog: Introducing Visual Edits",
      "url": "https://lovable.dev/blog/introducing-visual-edits",
      "published": "2025-02-12"
    }
  ],
  "image": {
    "url": "https://rangoon.ai/assets/editorial/archive-lovable-2-april-2025-diagram-v1.webp",
    "thumbnail_url": "https://rangoon.ai/assets/editorial/archive-lovable-2-april-2025-diagram-v1-thumb.webp",
    "alt": "Original build explanatory diagram, Team inputs through build modes. Labels: Workspace, Chat Mode, Dev Mode, Visual edits, Security scan, Publish review. Source boundary: dated primary release reporting; original editorial analysis.",
    "type": "original-diagram",
    "caption": "Team inputs through build modes. Original Rangoon explanatory diagram: Source boundary: dated primary release reporting; original editorial analysis. No vendor interface, photograph, measurement result, or endorsement implied.",
    "credit": "Rangoon / Hypler",
    "source_url": "https://rangoon.ai/brand/",
    "rights_url": "https://rangoon.ai/terms/#media-use",
    "rights_note": "Hypler artwork permissions do not grant rights to third-party marks or imply endorsement."
  },
  "sections": [
    {
      "id": "two-point-oh",
      "url": "https://rangoon.ai/insights/archive-lovable-2-april-2025/#two-point-oh",
      "heading": "The release bundled product collaboration with agent controls",
      "paragraphs": [
        "Lovable announced Lovable 2.0 on April 24, 2025 as a product update for people building apps and websites through AI chat. The dated release grouped a new brand and user interface with workspaces, multiplayer collaboration, a Chat Mode Agent, Security Scan, Dev Mode, Visual Edits and custom domains. This was a product-surface event rather than a new model release. Its technical significance lies in the separation of several work modes: users could collaborate around a project, ask an agent to inspect or plan, edit code directly, or make visual changes. The announcement does not say those modes shared the same permissions, review trail or safety guarantees. A responsible team would still decide which actions may be delegated and who owns approval for changes that affect users or data."
      ],
      "bullets": [],
      "source_ids": [
        1
      ]
    },
    {
      "id": "workspace-roles",
      "url": "https://rangoon.ai/insights/archive-lovable-2-april-2025/#workspace-roles",
      "heading": "Collaboration carried plan-specific roles and shared credits",
      "paragraphs": [
        "The workspace section of the release is more specific than a generic multiplayer claim. Lovable said each subscription was connected to a workspace. Pro subscribers had personal workspaces and could invite unlimited collaborators to individual projects, using the project owner’s credits. Teams subscribers could have up to 20 users in a workspace with a shared credit pool. The release also distinguished owner and administrator powers from editor powers: owners and administrators could invite or remove users, while editors managed and edited existing projects. Those statements describe the announced collaboration model, not a complete access-control audit. They do not establish whether a specific organization had the feature, how later role definitions changed, or whether project-level roles were sufficient for a particular source repository, database or deployment environment."
      ],
      "bullets": [],
      "source_ids": [
        1
      ]
    },
    {
      "id": "chat-and-code",
      "url": "https://rangoon.ai/insights/archive-lovable-2-april-2025/#chat-and-code",
      "heading": "Planning without edits sat beside direct source editing",
      "paragraphs": [
        "Lovable described its new Chat Mode Agent as a mode that did not make code edits. The release said it could ask questions, plan a project and debug, while deciding when to search files, inspect logs or query a database across multiple steps. That boundary is useful because an exploratory or diagnostic exchange should not silently become a source mutation. The same page described Dev Mode as a way to edit project code directly. Its Visual Edits reference can be read with Lovable’s earlier, dated February announcement, which described selecting visual styles and changing them directly. Together those records show a layered tool surface: chat for reasoning, code for explicit source edits and visual controls for presentation changes. They do not prove that a mode correctly understands system context or preserves design and code quality."
      ],
      "bullets": [],
      "source_ids": [
        1,
        2
      ]
    },
    {
      "id": "security-boundary",
      "url": "https://rangoon.ai/insights/archive-lovable-2-april-2025/#security-boundary",
      "heading": "A publish-time scan was a signal, not a complete security decision",
      "paragraphs": [
        "Lovable said its Security Scan could surface security vulnerabilities and appeared when a user clicked publish in a project connected to Supabase. The qualifier is central: the feature was tied to a particular integration and was described as a scan, not as a certification that the application was secure. A scan can find a class of issues while missing insecure authorization, data exposure, third-party integration failures, flawed business logic, vulnerable dependencies or unsafe operational practices. The release also mentioned built-in custom-domain buying and connection, which adds a separate ownership and deployment boundary outside the design conversation. The archival claim is limited to what Lovable announced: a new collaboration and agent workflow with a named scan. Deployment approval still requires independent review of code, policies, database access, secrets and runtime behavior."
      ],
      "bullets": [],
      "source_ids": [
        1
      ]
    }
  ],
  "project_context": null,
  "related_links": [],
  "permissions_url": "https://rangoon.ai/permissions.txt",
  "content_hash": "sha256:834decc7d32d3fc0345a7753b81ab005c1d6d6228da7a665525a270f9299592b"
}