Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@agent-kb how do I deploy the project?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
A drop-in, local-first knowledge base for LLM coding agents. Index your repository's documentation, concept ontology, and build targets into Qdrant, and expose retrieval to any MCP-capable agent (Claude Code, Cursor, custom agents) as tools — not context stuffing.
No closed corpora, no paid APIs, no managed vector database: one Qdrant container, CPU-local embeddings via fastembed, and a small Python MCP server.
Agents lose accuracy the moment they're asked about anything outside their pre-training corpus. The usual fix is RAG, but how the agent consumes retrieval matters more than the embedding model: inline context-stuffing burns tokens and amplifies noise; tool-shaped retrieval lets the agent ground claims selectively, the same way it already uses other tools.
Grounding splits into three layers with different staleness profiles:
Layer | Question shape | Owned by |
Prose | "How does X work?" "What's the deployment story?" | this KB (markdown sources) |
Ontology | "Which class implements concept X?" | this KB (concept → symbol bindings) |
Live code | "Where is that class right now? Who calls it?" | your LSP — see agent-code-intel |
The KB deliberately stops at the symbol name. It never stores file paths or line numbers for code — that's the LSP's job, and indexing source into a vector store just guarantees churn.
Related MCP server: agentmako
The repo doubles as a plugin marketplace covering both grounding repos:
/plugin marketplace add zmij/agent-kb
/plugin install agent-kb@agent-grounding
/plugin install agent-code-intel@agent-grounding # optional: the LSP layerThe agent-kb plugin ships the knowledge-base operating skill and the
MCP server — Claude Code launches kb serve-mcp via uv run from the plugin
checkout, and the server discovers your repo root by walking up from the
session's working directory to the nearest kb.yaml. No per-worktree
registration needed.
Then type /kb-setup in your repo: the bundled setup skill walks the agent
through the rest — start Qdrant, author a starter kb.yaml (it asks which
doc trees to index), run the first index, and verify search. Requires
uv on PATH and Docker for Qdrant.
Requirements: Python 3.11+, uv, Docker (for Qdrant).
git clone https://github.com/zmij/agent-kb # next to your repo, or as a submodule
cd agent-kb
make install # uv venv + editable install
make up # start Qdrant (docker compose)Or from your own repo, if you vendor this as a submodule and include kb.mk
in your Makefile (see Make integration):
make kb-install kb-upCreate kb.yaml at your repository's root (not in this repo). It declares
the project identity and the sources to index:
project: my-project # → collection "my_project_kb", MCP server "my-project-kb"
sources:
user_docs:
type: markdown # heading-aware chunking, frontmatter lifted to payload
root: docs/guides
uri_prefix: "docs://guides" # optional: preserve your internal link scheme
arch_docs:
type: markdown
root: docs
exclude: [guides, ontology] # subtrees that have their own indexers
ontology:
type: ontology # concept → code-symbol bindings (see below)
root: docs/ontology
make_targets:
type: make_targets # every documented `target: ## description`
files: [Makefile, "scripts/make/*.mk"]
# Only needed if you use the ontology maintenance loop (verify/heal/suggest-new)
symbols:
language: cpp
include_root: include/myproject # public header tree to parse
base_classes: [Strategy] # subclasses of these are discoverable concepts
strip_suffixes: [Strategy, Impl] # trimmed when deriving stub titles/slugs
stub_subdir: strategies # stubs land in docs/ontology/strategies/
stub_kind: strategy # `kind:` value written into stubsAll sources share one collection by default (cross-source retrieval in a
single search); override per-source with collection: or globally with
default_collection:.
kb index --all # chunk → embed → upsert (incremental by default)
kb search "how do I deploy" # semantic search across all sources
kb sources # what's indexed, per source
kb index user_docs --full # re-embed one source from scratchIncremental indexing hashes file content and only re-embeds changed files; chunks of deleted files are evicted automatically.
kb serve-mcp # stdio MCP serverFor Claude Code, register per project/worktree:
claude mcp add my-project-kb -e KB_REPO_ROOT=$(pwd) -- \
/path/to/agent-kb/.venv/bin/kb serve-mcp(the kb-register make target below does this for you, self-healing).
Tools exposed: kb_search, kb_get, kb_list_sources, kb_reindex.
An ontology entry is a small markdown file binding a domain concept to the code symbols that implement it:
---
concept: x-wing
title: X-Wing
kind: technique
implements:
- sudoku::XWingTechnique
related_concepts: [swordfish]
---
A fish pattern on two rows and two columns…The chunk text bakes the symbol list into the embeddable body, so "which class implements X-Wing" hits the bound names, not just prose. And because bindings are curated, they need a maintenance loop:
Command | What it does |
| Checks every |
| Proposes replacements for broken bindings by name-similarity against current symbols (deterministic, stdlib-only). |
| Finds subclasses of |
Currently the symbol parser covers C++ headers (tree-sitter). Other languages:
PRs welcome — the parser interface is one function, parse_header(path, root) -> [Symbol].
kb.mk ships includable targets (kb-up, kb-index, kb-search,
kb-register, kb-verify, …). From a consuming repo:
KB_DIR := tools/knowledge_base # wherever the submodule/clone lives
KB_MCP_NAME := my-project-kb
include $(KB_DIR)/kb.mkRun make kb-help for the full target list.
If you use git worktrees, each worktree writes to its own Qdrant collections,
suffixed with a slug derived from the worktree directory name
(my_project_kb_backend, my_project_kb_frontend, …). Concurrent indexing
across worktrees never collides, while one Qdrant container serves them all.
The slug derives from KB_REPO_ROOT — kb-register bakes
KB_REPO_ROOT=$(pwd) into the MCP registration so queries always land in the
registering worktree's collections. kb collections --all shows every
worktree's collections.
Environment (infrastructure — machine-level, .env supported):
Variable | Default | Purpose |
| auto-detected (kb.yaml walk-up) | Consuming repo root |
|
| Project config path |
| derived from repo root basename | Collection suffix override |
|
|
|
|
| Embedding model |
|
| Qdrant endpoint |
Switching embedding backends changes vector dimensions — re-index with
kb index --all --full.
agent-kb/
├── kb.example.yaml # annotated project-config template
├── kb.mk # includable make module
├── docker-compose.yml # Qdrant
├── src/kb/
│ ├── config.py # Settings (env) + KBConfig (kb.yaml)
│ ├── chunking/ # heading-aware markdown chunker
│ ├── embedding/ # provider Protocol + fastembed/ollama backends
│ ├── indexers/ # markdown / ontology / make_targets
│ ├── parsing/ # C++ header parser (tree-sitter)
│ ├── runner.py # chunk → embed → upsert, incremental + retries
│ ├── verify.py, heal.py, discover.py # ontology maintenance loop
│ ├── qdrant_client.py # store wrapper, per-worktree namespacing
│ ├── mcp_server.py # stdio MCP server
│ └── cli.py # `kb` entrypoint
├── tests/
└── docs/WORKSHOP.md # background: design rationale and the workshop storyIf you are an LLM agent working in a repo that uses agent-kb, read AGENTS.md for when to search the KB, how to phrase queries per source, and how to run the ontology maintenance loop.
main is branch-protected: no direct pushes (admins and their agents
included), linear history, everything lands through a pull request. Run
make test before opening one. This is the same Gate discipline the
tool exists to enforce — the repo practises it on itself.
MIT.
Fetch the full text and payload of a chunk by id.
Args:
id: The chunk id returned by kb_search.
source: Optional source hint to scope which collection to look in.
If omitted, every known collection is checked.
| Name | Required | Description | Default |
|---|---|---|---|
| id | Yes | ||
| source | No |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full behavioral disclosure burden. It explains the result content ('full text and payload') and adds important scoping behavior: when source is omitted, every known collection is checked. It doesn't cover not-found behavior or auth requirements, but for a simple getter this is reasonably transparent.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is compact and front-loaded: one clear summary sentence followed by a concise Args section. Every sentence contributes meaningful information with no filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a two-parameter getter with an output schema present, the description covers the essential call semantics and parameter behavior. It could have explicitly pointed to kb_list_sources for valid source values or described not-found behavior, but these are minor gaps given the tool's simplicity and the availability of an output schema.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, and the description fully compensates by explaining both parameters. It defines 'id' as the chunk id returned by kb_search, and 'source' as an optional collection-scoping hint with explicit default behavior when omitted. This adds real meaning beyond the raw schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description opens with a specific verb and resource: 'Fetch the full text and payload of a chunk by id.' This clearly distinguishes the tool from siblings like kb_search, kb_list_sources, and kb_reindex, which serve different operations.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives clear context: the id should come from 'kb_search', and the source hint scopes the lookup, with the default behavior of checking every known collection if omitted. It doesn't explicitly state when not to use this tool or name alternatives, but the linkage to kb_search provides strong usage guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List configured sources and the state of their Qdrant collections.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description must carry the behavioral disclosure burden. 'List' implies a read-only operation and 'state of their Qdrant collections' indicates the kind of information returned. However, it does not explicitly state that it is non-mutating, requires no setup, or has no side effects. It is minimally adequate but not rich.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, direct sentence that efficiently communicates the tool's purpose. Every word earns its place, and the core action is front-loaded with 'List configured sources.' No filler or redundant phrasing is present.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool has zero parameters and an output schema is present, the description does not need to explain return values. It captures the essential scope ('configured sources' and collection state). It could be more complete by noting how this relates to sibling tools, such as using it before reindexing, but overall it is sufficient for a simple list operation.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool has zero parameters, so schema coverage is trivially 100%. The baseline for 0-parameter tools is 4. The description does not add parameter-specific meaning, but none is needed because there are no parameters.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states a specific verb ('List'), a specific resource ('configured sources'), and the additional detail of 'the state of their Qdrant collections.' This distinguishes it from the sibling tools (kb_search, kb_get, kb_reindex), which imply searching, retrieving, and reindexing rather than listing sources.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives no explicit guidance on when to use this tool versus the siblings. It does not mention scenarios like checking configured sources, verifying collection state before reindexing, or troubleshooting. Usage must be inferred from the word 'List,' so this is a clear gap.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Re-run an indexer.
Args:
source: Indexer name from kb_list_sources.
| Name | Required | Description | Default |
|---|---|---|---|
| source | Yes |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description must disclose behavioral traits itself. It only says 'Re-run an indexer' and does not mention side effects, runtime, permissions, idempotency, or potential impact, leaving the agent with little safety context.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is extremely lean: one action sentence plus one parameter explanation. Every element carries information, and the core purpose is front-loaded.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a one-parameter tool with an output schema, the core invocation details are present. However, with zero annotations, the absence of any behavioral context such as side effects or prerequisites leaves moderate gaps for an agent deciding whether and how to invoke it.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema only defines source as a string, and schema description coverage is 0%. The description compensates by explaining that source is an indexer name and tells the caller to obtain it from kb_list_sources, which is meaningful guidance.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states a clear verb and resource: 'Re-run an indexer.' This is distinct from the sibling tools kb_search, kb_get, and kb_list_sources, though it does not elaborate on what re-running entails.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It implies the intended workflow by directing the caller to use an indexer name from kb_list_sources, but it does not explicitly state when to use this tool over alternatives or when not to use it.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Semantic search across indexed knowledge.
Args:
query: Natural-language question or keyword phrase.
source: Optional source name (see kb_list_sources) to filter
results. If omitted, searches every configured collection.
top_k: Maximum hits to return per collection.
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | ||
| top_k | No | ||
| source | No |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description bears the burden of explaining behavior. It adds non-obvious details: omitting source searches every configured collection, and top_k is per collection. It does not mention read-only status, auth, or rate limits, but search inherently implies a non-mutating operation and the output schema covers result structure.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is compact and well-structured: a brief one-line summary followed by a clean Args block. Every sentence and bullet adds necessary semantic value, with no filler or repetition.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The description fully covers invocation details for all parameters and references kb_list_sources for source names. The existence of an output schema means return values do not need to be described. The only minor gap is the lack of explicit guidance on choosing this tool versus kb_get or kb_reindex, but this does not prevent correct invocation.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema has 0% description coverage, yet the tool description fully explains all three parameters. It clarifies that query is a natural-language question or keyword phrase, source is optional and filterable via kb_list_sources, and top_k limits hits per collection. This is excellent compensation for the schema's lack of descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description opens with 'Semantic search across indexed knowledge,' which clearly names the operation (semantic search) and the resource (indexed knowledge). It readily distinguishes this tool from siblings like kb_get, kb_list_sources, and kb_reindex, which naturally serve different operations.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives useful parameter-level guidance, such as using an optional source and referencing kb_list_sources, but it does not explicitly state when to prefer kb_search over kb_get or kb_reindex. The intended usage is implied by the name and opening sentence, but no exclusions or alternative tool routing are provided.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Each tool targets a distinct operation: search, fetch-by-id, list sources, and reindex. There is no meaningful overlap; kb_search and kb_get are clearly separated by query-based retrieval versus direct ID lookup.
All tools use the consistent kb_ prefix followed by a clear verb or verb-noun phrase: search, get, list_sources, reindex. The naming pattern is uniform and predictable.
Four tools is compact but well-scoped for a knowledge base server covering retrieval, inspection, source listing, and maintenance. Each tool earns its place without unnecessary redundancy.
Core knowledge base operations are covered: search, fetch chunk content, list sources, and reindex. Minor gaps exist such as no delete/clear collection or detailed status beyond source listing, but these are not critical for the apparent purpose.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Self-hosted AI-native knowledge workspace with hybrid search, GraphRAG, and MCP.
Shared memory for coding agents. Stop re-explaining your codebase every session.
Knowledge base MCP for AI agents on iknow.dev. Search, read, and maintain via OAuth.
Shared knowledge base for AI agents. Semantic search across agents, no setup required — just a URL.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/zmij/agent-kb'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
