AI agent on-demand access to private project docs via BM25 ranked search.
Your AI agent doesn't know your project. Alcove fixes that.
English · 한국어 · 日本語 · 简体中文 · Español · हिन्दी · Português · Deutsch · Français · Русский
Alcove is an MCP server that gives AI coding agents on-demand access to your private project docs — without dumping everything into the context window, without leaking docs into public repos, and without per-project config for every agent you use.
Demo
Claude, Gemini, Codex — search · switch projects · global search · validate & generate. One setup.
CLI demo
alcove search· project switch ·--scope global·alcove validate
The problem
You have two bad options.
Option A: Put docs in CLAUDE.md / AGENTS.md
Every file gets injected into the context window on every run.
Works for short conventions. Breaks down with real project docs.
10 architecture files = context bloat = slower, more expensive, less accurate responses.
Option B: Don't put docs in Your agent invents requirements you already documented. It ignores constraints from decisions you already made. It asks you to explain the same things every session.
Neither option scales. Now multiply it across 5 projects and 3 agents, each configured differently. Every time you switch, you lose context.
How Alcove solves this
Alcove doesn't inject your docs. Agents search for what they need, when they need it.
~/projects/my-app $ claude "how is auth implemented?"
→ Alcove detects project: my-app
→ BM25 search: "auth" → ARCHITECTURE.md (score: 0.94), DECISIONS.md (score: 0.71)
→ Agent gets the 2 most relevant docs, not all 12
~/projects/my-api $ codex "review the API design"
→ Alcove detects project: my-api
→ Same doc structure, same access pattern
→ Different project, zero reconfiguration
Switch agents anytime. Switch projects anytime. The document layer stays standardized.
The right split
CLAUDE.md / AGENTS.md is for agent behavior: repeated mistakes to avoid, coding conventions, and session-specific instructions. Keep it under 200 lines.
Alcove is for project knowledge: architecture, decisions, runbooks, API docs, and anything else your agent needs to understand — but not necessarily on every run.
The pattern:
CLAUDE.md | AGENTS.md ← agent rules, coding conventions, recurring corrections
~/.config/alcove/docs/my-app/
ARCHITECTURE.md ← tech stack, data model, system design
DECISIONS.md ← why X was chosen over Y
DEBT.md ← known issues, workarounds
... ← agent searches here when it needs context
Agents call search_project_docs("auth flow") and get the 2 most relevant docs — not all 12. Nothing hits the context window unless it's actually needed.
Why Alcove
Why not just use
CLAUDE.md? Short conventions and agent behaviors belong there. Project documentation — architecture, decisions, runbooks, PRDs — doesn't scale in a context file. Alcove is not a replacement; it's the layerCLAUDE.mdwas never meant to be.
| Without Alcove | With Alcove |
|---|---|
Docs in CLAUDE.md bloat context on every run |
BM25 search — agents pull only what they need |
| Internal docs scattered across Notion, Google Docs, local files | One doc-repo, structured by project |
| Each AI agent configured separately for doc access | One setup, all agents share the same access |
| Switching projects means re-explaining context | CWD auto-detection, instant project switch |
| Agent search returns random matching lines | Ranked results — best matches first, one result per file |
| "Search all my notes about OAuth" — impossible | Global search across every project in one query |
| Sensitive docs sitting in project repos | Pr |
Tools (1)
search_project_docsSearches project documentation using BM25 ranking to retrieve the most relevant context for the agent.Configuration
{"mcpServers": {"alcove": {"command": "alcove", "args": ["--path", "/path/to/your/docs"]}}}