{"id":50391008,"url":"https://github.com/hallelx2/vectorless-mcp","last_synced_at":"2026-05-30T18:01:35.558Z","repository":{"id":353866849,"uuid":"1219985340","full_name":"hallelx2/vectorless-mcp","owner":"hallelx2","description":"MCP server for Vectorless — give AI agents access to structure-preserving document retrieval","archived":false,"fork":false,"pushed_at":"2026-04-24T12:52:11.000Z","size":33,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-04-26T01:18:02.140Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/hallelx2.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-04-24T12:29:34.000Z","updated_at":"2026-04-24T12:52:10.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/hallelx2/vectorless-mcp","commit_stats":null,"previous_names":["hallelx2/vectorless-mcp"],"tags_count":2,"template":false,"template_full_name":null,"purl":"pkg:github/hallelx2/vectorless-mcp","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hallelx2%2Fvectorless-mcp","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hallelx2%2Fvectorless-mcp/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hallelx2%2Fvectorless-mcp/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hallelx2%2Fvectorless-mcp/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/hallelx2","download_url":"https://codeload.github.com/hallelx2/vectorless-mcp/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hallelx2%2Fvectorless-mcp/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":33703065,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-05-30T02:00:06.278Z","response_time":92,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2026-05-30T18:01:34.674Z","updated_at":"2026-05-30T18:01:35.552Z","avatar_url":"https://github.com/hallelx2.png","language":"TypeScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003ch1 align=\"center\"\u003e\n  \u003ccode\u003evectorless-mcp\u003c/code\u003e\n\u003c/h1\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003cstrong\u003eMCP server for Vectorless — give AI agents access to structure-preserving document retrieval.\u003c/strong\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://www.npmjs.com/package/vectorless-mcp\"\u003e\u003cimg src=\"https://img.shields.io/npm/v/vectorless-mcp?style=flat-square\u0026logo=npm\u0026logoColor=white\u0026color=CB3837\" alt=\"npm\" /\u003e\u003c/a\u003e\n  \u003ca href=\"https://modelcontextprotocol.io\"\u003e\u003cimg src=\"https://img.shields.io/badge/MCP-compatible-4A90D9?style=flat-square\" alt=\"MCP\" /\u003e\u003c/a\u003e\n  \u003cimg src=\"https://img.shields.io/badge/OAuth_2.1-PKCE-green?style=flat-square\" alt=\"OAuth\" /\u003e\n  \u003cimg src=\"https://img.shields.io/badge/transport-Streamable_HTTP-blue?style=flat-square\" alt=\"Streamable HTTP\" /\u003e\n  \u003ca href=\"LICENSE\"\u003e\u003cimg src=\"https://img.shields.io/badge/license-MIT-blue?style=flat-square\" alt=\"License\" /\u003e\u003c/a\u003e\n\u003c/p\u003e\n\n---\n\n## How It Works\n\nVectorless provides a **remote MCP server** at `https://api.vectorless.store/mcp`. Your AI assistant connects to it directly over the internet — no local process, no API keys to manage.\n\nAuthentication uses **OAuth 2.1 with PKCE**: on first connect your AI client opens the Vectorless dashboard where you log in and approve access. After that, it just works.\n\n```\n┌──────────────────────────────────────────────────────────────┐\n│        Claude Desktop / Cursor / Windsurf / Claude Code      │\n│                                                              │\n│   1. Add the remote MCP URL                                  │\n│   2. OAuth redirect → Vectorless Dashboard (login + consent) │\n│   3. Token issued → MCP session active                       │\n│   4. AI calls tools via Streamable HTTP + SSE                │\n└───────────────────────────┬──────────────────────────────────┘\n                            │\n                            │  Streamable HTTP + SSE\n                            ▼\n┌──────────────────────────────────────────────────────────────┐\n│                  api.vectorless.store/mcp                     │\n│                                                              │\n│   OAuth 2.1 (PKCE)          7 Tools                          │\n│   ┌──────────────────┐      ┌──────────────────────────┐    │\n│   │ /.well-known/    │      │ vectorless_list_documents │    │\n│   │   oauth-*        │      │ vectorless_ingest_document│    │\n│   │ /oauth/authorize │      │ vectorless_get_document   │    │\n│   │ /oauth/token     │      │ vectorless_get_tree       │    │\n│   │ /oauth/register  │      │ vectorless_get_section    │    │\n│   └──────────────────┘      │ vectorless_query          │    │\n│                              │ vectorless_delete_document│    │\n│   Scopes:                    └──────────────────────────┘    │\n│   • documents:read                                           │\n│   • documents:write          Vectorless Engine               │\n│   • query                    (retrieval + ingestion)         │\n└──────────────────────────────────────────────────────────────┘\n```\n\n## Setup (Remote — Recommended)\n\nAdd the remote URL to your AI client config. **No installation needed.**\n\n### Claude Desktop\n\n`Settings → Developer → Edit Config`:\n\n```json\n{\n  \"mcpServers\": {\n    \"vectorless\": {\n      \"url\": \"https://api.vectorless.store/mcp\"\n    }\n  }\n}\n```\n\n### Cursor\n\n`.cursor/mcp.json`:\n\n```json\n{\n  \"mcpServers\": {\n    \"vectorless\": {\n      \"url\": \"https://api.vectorless.store/mcp\"\n    }\n  }\n}\n```\n\n### Windsurf\n\n`~/.codeium/windsurf/mcp_config.json`:\n\n```json\n{\n  \"mcpServers\": {\n    \"vectorless\": {\n      \"url\": \"https://api.vectorless.store/mcp\"\n    }\n  }\n}\n```\n\n### Claude Code\n\n`.claude/settings.json`:\n\n```json\n{\n  \"mcpServers\": {\n    \"vectorless\": {\n      \"url\": \"https://api.vectorless.store/mcp\"\n    }\n  }\n}\n```\n\nThat's it. On first use, your AI client opens the Vectorless dashboard in your browser. Log in (Google, GitHub, or email), approve the permissions, and you're connected.\n\n## OAuth Scopes\n\nWhen you authorize, the consent screen shows these permissions:\n\n| Scope | Grants access to |\n|-------|-----------------|\n| **`documents:read`** | List documents, get metadata, browse tree, fetch sections |\n| **`documents:write`** | Upload new documents, delete existing ones |\n| **`query`** | Run natural language queries against documents |\n\nYou choose which scopes to grant. The AI agent can only use tools matching your approved scopes.\n\n## Available Tools\n\n### `vectorless_list_documents`\nList all your documents with status, metadata, and pagination.\n\n### `vectorless_ingest_document`\nUpload a document from a URL or base64 content. Supports PDF, DOCX, Markdown, HTML, plain text. Documents are parsed into hierarchical section trees with AI-generated summaries.\n\n### `vectorless_get_document`\nGet metadata and processing status for a document (pending → parsing → summarizing → ready).\n\n### `vectorless_get_tree`\nView the hierarchical outline — sections with titles, summaries, depth, and token counts. Use this to understand what's in a document before querying.\n\n### `vectorless_get_section`\nFetch the full content of a specific section by ID.\n\n### `vectorless_query`\nAsk a natural language question. An LLM navigates the document tree to find and return the most relevant sections with full content, strategy, timing, and cost.\n\n### `vectorless_delete_document`\nPermanently delete a document and all its sections.\n\n## Example Conversation\n\n```\nYou:    Upload this paper: https://arxiv.org/pdf/2301.00001\nClaude: [calls vectorless_ingest_document]\n        Done — doc_abc123 is ready (14 sections, 48,200 tokens).\n\nYou:    What's the structure?\nClaude: [calls vectorless_get_tree]\n        Introduction (1,200 tokens)\n          Background (800 tokens)\n          Related Work (2,400 tokens)\n        Methodology (3,600 tokens)\n          Data Collection (1,100 tokens)\n          Model Architecture (2,500 tokens)\n        Results (4,200 tokens)\n        Conclusion (900 tokens)\n\nYou:    What model architecture did they use?\nClaude: [calls vectorless_query]\n        Found 2 relevant sections (chunked-tree, 340ms):\n\n        ## Model Architecture\n        We employ a transformer-based encoder with 12 attention heads...\n\n        ## Results\n        The model achieved 94.2% accuracy on the benchmark...\n```\n\n## Fallback: Local Stdio Server\n\nIf your MCP client doesn't support remote servers, run the stdio bridge locally:\n\n```bash\nnpx vectorless-mcp\n```\n\n```json\n{\n  \"mcpServers\": {\n    \"vectorless\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"vectorless-mcp\"],\n      \"env\": {\n        \"VECTORLESS_API_KEY\": \"vl_...\"\n      }\n    }\n  }\n}\n```\n\n| Variable | Default | Description |\n|----------|---------|-------------|\n| `VECTORLESS_API_KEY` | — | API key (required for remote, optional for self-hosted) |\n| `VECTORLESS_BASE_URL` | `https://api.vectorless.store` | Server URL |\n| `VECTORLESS_TRANSPORT` | `http` | Wire protocol (`http` or `connect`) |\n\n## Self-Hosted\n\nPoint the MCP to your own Vectorless instance:\n\n```json\n{\n  \"mcpServers\": {\n    \"vectorless\": {\n      \"url\": \"https://your-server.com/mcp\"\n    }\n  }\n}\n```\n\nOr via stdio:\n\n```json\n{\n  \"mcpServers\": {\n    \"vectorless\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"vectorless-mcp\"],\n      \"env\": {\n        \"VECTORLESS_BASE_URL\": \"https://your-server.com\",\n        \"VECTORLESS_API_KEY\": \"your-key\"\n      }\n    }\n  }\n}\n```\n\n## License\n\nMIT\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhallelx2%2Fvectorless-mcp","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fhallelx2%2Fvectorless-mcp","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhallelx2%2Fvectorless-mcp/lists"}