AKB — Agent Knowledge Base

Organizational memory for AI agents. Git-backed knowledge base served over the Model Context Protocol (MCP) — agents read and write directly with hybrid semantic + keyword search, structured tables, files, and a URI graph. Drop-in alternative to Confluence / Notion for Claude Code, Cursor, Windsurf, and any MCP-aware agent.

Works with

Any agent client that speaks MCP (Streamable HTTP or stdio):

• Claude Code — CLI / VS Code / JetBrains
• Claude Desktop — macOS / Windows
• Cursor, Windsurf, Cline, Continue — via the akb-mcp stdio proxy
• Custom agents — direct HTTP POST /mcp/ with a Bearer token

The default flow uses a Personal Access Token. Deployments with the optional MCP OAuth Resource Server path turned on (via Keycloak as the AS — see docs/mcp-clients/web-connectors.md) also accept Claude Code's mcp add --transport http + mcp login flow end-to-end, without a PAT.

Plugins

Beyond raw MCP access, AKB ships ready-made agent plugins for Claude Code and Codex that wrap common vault workflows:

• akb-wiki — ingest a source (local file, web URL, GitHub PR / release / commit, Confluence page, or Jira issue) into the vault as a structured document, and answer questions from the vault with grounded, cited synthesis (read-only).
• akb-sessions — capture a coding session as structured notes: a session report plus follow-up tasks, learnings, ideas, and decisions.
• akb-claude-code — a Claude Code lifecycle bridge: hooks anchor each session to your AKB memory vault, injecting preferences and recent learnings at the start and writing a recap at the end.

/plugin marketplace add dnotitia/akb        # Claude Code
codex plugin marketplace add dnotitia/akb   # Codex

Install details and credentials: plugins/.

Try it live

A public demo runs at akb-demo.agent.seahorse.dnotitia.ai. Browse and search a small fictional-organization knowledge base — product docs, a company handbook, agent session notes, and an engineering wiki, cross-linked by the URI graph — right in your browser, no signup. To wire it into your own agent, sign up with any email (a throwaway address is fine) and point the akb-mcp proxy at https://akb-demo.agent.seahorse.dnotitia.ai/mcp/.

⚠️ Throwaway demo. It is public, wiped and re-seeded weekly, and runs on minimal resources with no uptime, privacy, or data guarantees. Don't put anything real or sensitive in it — treat every write as public and ephemeral. For real use, self-host in three containers.

Why AKB

Most knowledge tools are built for humans clicking through a UI. Agents need a different shape: structured documents, semantic + keyword search in one call, explicit relations, and full version history. AKB gives agents a single set of tools (akb_put, akb_search, akb_browse, akb_relations, …) over a backing store of Git bare repos and a PostgreSQL hybrid index.

Retrieval quality

Memory is only useful if the right note comes back. AKB's hybrid retrieval (dense + BM25, source-level dedup) was benchmarked on LongMemEval-S — 500 long-context questions, ~50 chat sessions per question. Recall@5 = 98.4%, with no reranker in the loop.

Methodology, per-category breakdown, and a one-command reproducible harness live in eval/longmemeval/. The embedding model differs across systems (AKB: bge-m3@1024), so read this as a stack-level comparison.

Design philosophy

Core stays small; flexibility comes from extension, not built-in automation. AKB does not ship its own consolidator, summariser, or "knowledge gardener" — instead every write emits a structured event to a Redis Stream (akb:events). Operators wire any external consumer (periodic synthesis bot, doc-rot reaper, weekly-digest agent, audit trail, …) on top, with no patches to the core. The base contract is a read/write store; opinions about what to do with the knowledge live outside.

Architecture

┌──────────────────────────────────────────────────────────┐
│                  Access Layer                            │
│   MCP Server  │  REST API  │  Web UI                     │
├──────────────────────────────────────────────────────────┤
│                  Core Services                           │
│   Document (Put/Get)  │  Search (Hybrid: dense+BM25)     │
│   Relations (graph)   │  Session  │  Publications        │
├──────────────────────────────────────────────────────────┤
│                  Storage Layer                           │
│   Git bare repos       │  PostgreSQL 16 (text + meta SoT)│
│                        │  Vector store (driver):         │
│                        │    pgvector        (default, PG)│
│                        │    qdrant          (optional)   │
│                        │    seahorse-cloud  (managed)    │
│                        │    seahorse-db     (self-hosted)│
│                        │    seahorse-db-grpc(experimental)│
└──────────────────────────────────────────────────────────┘

PostgreSQL is the source of truth — chunk text + metadata + BM25 vocab. The vector store is a driver-pluggable derived index holding dense embeddings and corpus-side sparse vectors. Full vector-store loss is recoverable from PG by setting chunks.vector_indexed_at = NULL and letting the indexing worker re-populate.

Key Concepts

• Vault — A Git bare repo. The unit of access control and physical isolation.
• Collection — A directory inside a vault. Topical grouping of documents.
• Document — Markdown + YAML frontmatter, optimised for agent read/write.
• Hybrid Search — Dense (semantic) + BM25 (lexical) fused via RRF in one call.
• Relations — depends_on, related_to, implements in frontmatter form an explicit knowledge graph.
• Vault isolation in akb_sql — Enforced by PostgreSQL ACL. Each AKB user has a corresponding PG role (akb_user_<uid>) and each vault has three group roles (akb_vault_<vid>_{reader,writer,admin}). akb_sql runs the user's SQL inside a transaction with SET LOCAL ROLE; cross-vault references return PG 42501 directly. No application-side regex inspects user SQL for forbidden identifiers. See docs/designs/pg-native-rbac/.

MCP Tools (selection)

Agent memory and session lifecycle are not MCP tools — they live on the dedicated /api/v1/agent-sessions REST surface, driven by AKB lifecycle plugins (akb-claude-code, akb-cursor, …) that hook into the agent's own SessionStart / PreCompact / SessionEnd events. As an agent, your own memory vault (agent-memory-{username}) is browsable through the standard akb_search / akb_browse / akb_get tools exactly like any other vault.

The full tool catalogue is exposed via akb_help() from any MCP client.

Inline document images from MCP

Inline images are hidden document attachments, not browsable Files. Upload a local PNG, JPEG, GIF, or WebP (maximum 10 MiB), then insert the returned Markdown without reconstructing its asset URL:

image = akb_put_image(
  parent="akb://eng/coll/specs",
  file_path="/workspace/architecture.png",
  alt_text="Request processing architecture")

akb_put(
  parent="akb://eng/coll/specs",
  title="Request Processing",
  content="# Architecture\n\n" + image.markdown)

For an existing document, use akb_get followed by a targeted akb_edit(base_commit=...). Do not pass only the image fragment to akb_update(content=...), which replaces the complete body. Image bytes are immutable: replacing an image means uploading a new one and editing the Markdown reference. Remove an image by deleting its Markdown expression; use akb_discard_image only for an upload that never reached a successful document commit. Run akb_help(topic="images") for retention and publication behavior.

The image tools require both the matching backend release and akb-mcp 2.2 or newer. For upgrades, deploy the backend first, then publish/install the proxy and restart existing MCP processes so they load the updated tool list.

Document Format

Every vault resource has a location-aware AKB URI — the canonical handle used by every tool and stored in relations. As of 0.3.0:

akb://{vault}                                          vault root (browse target)
akb://{vault}/coll/{coll_path}                         collection (browse target)
akb://{vault}[/coll/{coll_path}]/doc/{filename}        document
akb://{vault}[/coll/{coll_path}]/table/{name}          table
akb://{vault}[/coll/{coll_path}]/file/{uuid}           file

The /coll/{coll_path} segment is omitted for resources at the vault root. Walking up a URI to its parent collection is a pure string operation — paste the parent into akb_browse(uri=...) to list siblings without an extra lookup.

---
title: "Payment API v2 migration plan"
type: plan              # note | report | decision | spec | plan | session | task | reference
status: active          # draft | active | archived | superseded
tags: [payments, api]
domain: engineering
summary: "REST → gRPC transition plan."
depends_on: ["akb://eng/coll/specs/doc/payment-api-v2.md"]
related_to: ["akb://eng/coll/meetings/doc/2026-05-01-payments.md"]
---

# Payment API v2 migration plan
...

Open Knowledge Format (OKF) compatible

A vault is stored as a git tree of .md + YAML-frontmatter files whose identity is the path — the same model as Google Cloud's Open Knowledge Format (OKF v0.1), which AKB independently arrived at before the spec existed. AKB-authored bundles satisfy all three OKF MUST rules, and AKB can export any vault as a conformant OKF bundle (documents, plus tables/files as concept docs) and validate any bundle:

python -m app.cli okf-export --from-git /data/vaults/_worktrees/<vault> \
    --vault <vault> --out ./okf-out/
python -m app.cli okf-validate ./okf-out/

OKF and AKB are complementary — OKF standardizes how knowledge is written down; AKB stores, versions, searches, governs, and serves it to agents. See okf/ for the mapping and a sample bundle.

Quick Start

AKB ships as a 3-container stack (PostgreSQL with pgvector + backend + frontend). For semantic (dense) search you bring an OpenAI-compatible embedding endpoint (OpenAI, OpenRouter, self-hosted vLLM/TEI, etc.). It is not strictly required: with no embed endpoint (or during an outage) the pgvector and Qdrant drivers degrade to BM25-only lexical search rather than returning nothing — dense is genuinely optional end-to-end (the seahorse-db driver is the exception; see Vector store below). Prefer running a separate Qdrant cluster, or pointing at Seahorse? See Vector store below.

# 1. Configure
cp config/app.yaml.example   config/app.yaml
cp config/secret.yaml.example config/secret.yaml
$EDITOR config/secret.yaml   # set embed_api_key and replace system_hmac_secret

# Generate the installation's persistent RSA-3072 local-session keyset.
# This directory is gitignored; back it up with the other installation secrets.
cd backend
uv run python -m app.cli generate-local-session-keyset \
  --output-dir ../config/local-session
cd ..

# 2. Run
docker compose up -d

# 3. Provision the designated recovery administrator (local mode)
#    The password is read from stdin and is never printed by AKB.
docker compose exec -T backend python -m app.cli provision-recovery-admin local \
  --username recovery-admin --email [email protected] \
  --password-file - < /secure/operator/recovery-admin.password

# 4. Open
open http://localhost:3000

config/app.yaml and config/secret.yaml are the single source of runtime configuration — no environment variables are read by the backend. Mount the config/ directory at /etc/akb/ in any deployment.

Ordinary registration always creates a non-admin account, including on an empty database. Administrator bootstrap is available only through the operator CLI; there is no unauthenticated HTTP bootstrap endpoint. The CLI profile must match auth_mode:

# Local: have AKB generate the password only when an operator-owned output
# file is explicitly requested. A new file is created with mode 0600 and the
# password is not written to stdout, stderr, logs, or application config.
python -m app.cli provision-recovery-admin local \
  --username recovery-admin --email [email protected] \
  --generate-password-file /secure/operator/recovery-admin.password

# SSO: pre-bind the product administrator to the exact external identity.
# Username and email are snapshots; issuer + subject are the identity key.
python -m app.cli provision-recovery-admin sso \
  --username recovery-admin --email [email protected] \
  --issuer https://issuer.example.com/realms/akb \
  --subject exact-provider-subject

The same exact identity is idempotent. A different designation, an existing username/email, or an already-bound external identity fails closed. The SSO command stores no usable local password and does not contact the identity provider. Generated output files are create-only and never overwritten; for a retry after the file exists, pass that file back with --password-file.

Open /admin for the separate product-administration surface. In local mode it accepts the provisioned local administrator and returns the same local-session-rs256-v2 profile used by local human authentication, but it refuses non-admin accounts. In sso mode local credentials are absent: /admin uses a dedicated confidential akb-admin Keycloak client with PKCE and nonce, then accepts only the exact pre-bound (issuer, subject) whose AKB account is still active and is_admin=true.

In SSO mode the same /admin surface can configure a built-in upstream IdP, save it disabled, inspect its exact broker redirect URI, and enable or disable its ordinary-login option without redeploying AKB. The option becomes a usable button only when the server-side browser-session capability is ready. Client secrets are write-only, and an enabled provider must be disabled before reconfiguration. See the SSO provider guide and the first reference integration, Keycloak OIDC behind Keycloak.

The dedicated admin callback stores no Keycloak access, refresh, or ID token. It creates a short-lived opaque HttpOnly admin cookie plus a CSRF token; PostgreSQL stores only their hashes plus the exact identity snapshot, and rechecks the account, unchanged external binding, and admin flag on every request. Its one-time OIDC state is also bound to a short-lived HttpOnly cookie so a callback copied into another browser fails before token exchange. Configure keycloak_admin_client_secret, register <public_base_url>/api/v1/admin/auth/keycloak/callback and <public_base_url>/admin in the dedicated client, and keep the admin client ID out of every API/MCP resource-client path. Browser-facing AKB and Keycloak URLs must use HTTPS outside the explicit loopback development exception.

Ordinary SSO login uses the separate akb-web client. The browser receives only an opaque HttpOnly AKB session plus a readable CSRF value; SSO does not mint an AKB user JWT. AKB encrypts the Keycloak refresh/ID token set with the independent sso_browser_session_encryption_key and never persists an access token. The client must map Keycloak's identity_provider user-session note into both ID and access tokens with oidc-usersessionmodel-note-mapper; AKB binds that signed broker alias to the selected enabled provider on callback and every refresh. Production HTTPS cookies use the browser-enforced __Host- prefix, Secure, no Domain, and Path=/; loopback HTTP uses isolated development names. Generate the key as 32 random bytes encoded with unpadded base64url and keep it stable across restarts. See the Keycloak boundary for refresh, logout, and back-channel revocation details.

Local login issues only the versioned local-session-rs256-v2 profile: RS256 with an installation-owned RSA-3072 key, an RFC 7638 kid, exact deployment issuer/audience, jti, and a public-only JWKS at GET /api/v1/auth/jwks. AKB never chooses a verifier from an untrusted token alg header. Upgrading from an HS256 release is an intentional forced-login boundary: generate and persist the v2 keyset before rollout, set jwt_algorithm: RS256, and restart all backends together. Existing HS256 user sessions then receive 401 and must sign in again; PATs and service keys are not revoked. The old jwt_secret may be retained for one release only as migration input for short-lived internal HMAC capabilities, or renamed unchanged to system_hmac_secret; it is never accepted as human-session signing material.

For routine v2 key rotation, generate a new directory while retaining the current public JWKS, publish the new immutable Secret/config revision, and roll every backend to that exact pair:

cd backend
uv run python -m app.cli generate-local-session-keyset \
  --output-dir /secure/akb/local-session-next \
  --retain-jwks /secure/akb/local-session-current/jwks.json

Keep a retained public key for at least jwt_expire_hours plus rollout skew, then remove it in a later coordinated keyset revision. Restoring the previous private/JWKS pair is the rollback; never overwrite key files in place.

Vector store (driver-pluggable)

Hybrid search (dense + BM25 sparse, RRF-fused) runs through a driver interface. Five drivers ship; pick at config time:

• pgvector (default) — uses the same Postgres container that holds application data. The pgvector/pgvector image pre-installs the extension; the driver creates a separate vector_index schema, so the main chunks table stays plain PostgreSQL. RRF fusion runs application-side. No external service to operate.
• qdrant — runs a separate Qdrant container; native RRF via the Query API. Useful when you already operate Qdrant or want to scale the vector store independently of Postgres.
• seahorse-cloud — points at a managed Seahorse Cloud table over its BFF management API + per-table data-plane host (Bearer auth). No infrastructure to run on your side; you provision a table in the Seahorse console (or let the driver auto-create one) and AKB stores its chunks there. Native RRF, server-side BM25. See docs/vector-store-seahorse.md for the end-to-end setup walkthrough (sign-up → token → schema → config).
• seahorse-db — points at a self-hosted SeahorseDB cluster via its Coral coordinator HTTP API. You run Coral + Writer + Reader(s) + Redis + Kafka + a sparse-embedding server yourself (the SeahorseDB monorepo's deploy/docker-compose.yml brings up a minimal single-box stack). Native dense+sparse hybrid. Unlike the other drivers it does not support BM25-only fallback when the embed API is down (its sparse path is server-side and structurally coupled to a live embed step) — keep an embedding endpoint reachable for this driver.
• seahorse-db-grpc (experimental) — same Coral coordinator as seahorse-db, same seahorsedb_* settings, but talks gRPC instead of REST/JSONL. Coral merges axum + tonic onto a single listener so the port doesn't change; only the wire format does. Trades the JSON parsing path (and a class of foot-guns like INT64 sign mismatch and Arrow JSON decoder edge cases) for typed protobuf messages and an Arrow IPC streaming result. Prefer the REST driver for production until the gRPC variant clears its own QPS / recall benchmark. Same CRUD parity with REST (passes the same 25-scenario hybrid e2e), but it has not yet had the production-scale exposure the REST driver has.

Switching drivers is a config edit (no schema migration on the main DB):

# Default flow targets pgvector.
docker compose up

# Qdrant:
docker compose -f docker-compose.yaml -f docker-compose.qdrant.yaml up
$EDITOR config/app.yaml     # vector_store_driver: qdrant
                            # vector_url: http://qdrant:6333

# Seahorse Cloud (managed; full guide in docs/vector-store-seahorse.md):
docker compose up           # no extra container needed
$EDITOR config/app.yaml     # vector_store_driver: seahorse-cloud
                            # seahorse_cloud_tenant_uuid: <your tenant>
                            # seahorse_cloud_table_name: <your table>
$EDITOR config/secret.yaml  # seahorse_cloud_token: shsk_<...>

# SeahorseDB (self-hosted cluster reached via the Coral coordinator):
docker compose up           # run the SeahorseDB stack separately
$EDITOR config/app.yaml     # vector_store_driver: seahorse-db
                            # seahorsedb_coordinator_url: http://localhost:3003
                            # seahorsedb_table_name: akb_chunks

Embedding model + dimensions are also fully pluggable via embed_base_url / embed_model / embed_dimensions — the codebase has no hard-coded model. For pgvector with HNSW, keep embed_dimensions ≤ 2000 (or 4000 with halfvec); larger models fall back to exact scan. Qdrant / Seahorse (cloud or db) have no such limit (Qdrant up to 65536, Seahorse up to its table-defined dim).

LLM features (optional)

LLM is only used by the metadata_worker to auto-tag documents imported via external git mirroring. Core CRUD/search works without it. To enable, set llm_base_url / llm_model in app.yaml and llm_api_key in secret.yaml.

Standalone deployments default to model_api_governance_mode: external_metering and may point embedding, chat, and rerank at any compatible provider. A managed control plane can instead set platform_hard plus an exact platform_gateway_base_url. In that mode AKB fails startup if an active model route points anywhere else or lacks a credential, and every model call carries a caller-generated Idempotency-Key for durable gateway reservation/settlement. Gateway policy or budget denials are never fanned out into per-item retries.

Event fanout (optional)

The PG events outbox is always written. Set redis_url in app.yaml to have the events_publisher worker drain the outbox to a Redis Stream (akb:events) so external services can subscribe via XREAD / consumer groups. Leave blank to disable; events still accumulate in PG and you can build an SSE endpoint on top of the LISTEN/NOTIFY trigger without Redis.

Audit log (optional)

Off by default. Set audit.enabled: true in app.yaml to emit a structured, append-only, hash-chained JSON-lines audit log at the MCP dispatch chokepoint — every read, write, and auth denial, uniformly. AKB is a producer only: it does not store, query, or retain audit data; your SIEM (Splunk/QRadar/Elastic) scrapes the stream and owns retention under its own compliance regime. Each line carries a monotonic seq plus sha256(prev ‖ line), so the chain can be verified for dropped or altered lines and re-seeds from disk across restarts. Optionally hand the daily rolled file off to a WORM object-storage bucket (audit.bucket — provision with Object Lock and a write-only key for a true immutable trail); the local buffer is pruned only after a confirmed upload. Capture is best-effort and never raises into the serving path. See config/app.yaml.example for the full audit: block.

Production deployment

For Kubernetes, see deploy/k8s/README.md. The deploy/k8s/ directory contains a generic kustomize base; provide your own registry, hostname, and TLS issuer via the documented env vars or an operator-private overlay under deploy/k8s/internal/.

Project Structure

akb/
├── backend/                  # Python 3.14 / FastAPI / asyncpg / GitPython
│   ├── app/
│   │   ├── api/routes/       # REST endpoints
│   │   ├── services/         # Business logic + workers
│   │   └── db/               # PostgreSQL schema + migrations
│   ├── mcp_server/           # Streamable HTTP MCP server
│   └── tests/                # E2E shell tests
├── frontend/                 # React 19 + TypeScript + Vite + Tailwind
├── packages/
│   ├── akb-client/           # REST SDK boundary (npm: @akb/client)
│   └── akb-mcp-client/       # stdio ↔ HTTP MCP proxy (npm: akb-mcp)
├── agents/                   # Reference Python agent runtime (think/act loop over MCP)
├── plugins/                  # Claude Code / Codex agent plugins (ingest, query, session capture, lifecycle)
├── templates/                # Doc templates (ADR, PRD, runbook, …) and vault profiles
├── okf/                      # Open Knowledge Format interop: positioning + sample bundle
├── design-system/            # Frontend design system docs
├── config/
│   ├── app.yaml.example      # Non-secret runtime settings
│   └── secret.yaml.example   # API keys, passwords (gitignored when not .example)
├── deploy/
│   └── k8s/                  # Generic kustomize base for Kubernetes
└── docker-compose.yaml       # 3-container local stack (postgres + backend + frontend)

Tech Stack

• Backend: Python 3.14, FastAPI, Uvicorn, asyncpg, GitPython, MCP SDK
• Database: PostgreSQL 16 (main DB needs no extension; the same pgvector/pgvector image hosts the optional vector_index schema)
• Vector store: driver-pluggable (pgvector default; Qdrant, Seahorse Cloud, or self-hosted SeahorseDB optional — hybrid dense + BM25 sparse, RRF fusion; BM25-only fallback when embed is down)
• Event stream (optional): PG events outbox + Redis Streams fanout
• Audit log (optional): hash-chained append-only JSONL at the MCP dispatch point + optional WORM S3 handoff; producer-only (SIEM owns retention)
• Frontend: React 19, TypeScript, Vite, Tailwind CSS v4, Radix UI
• Auth: JWT + Personal Access Tokens (PATs)
• MCP: Streamable HTTP (backend) + stdio proxy (akb-mcp on npm)

Versioning

AKB follows SemVer. The product version lives in backend/pyproject.toml ([project].version) and is mirrored to frontend/package.json via scripts/bump-version.sh <x.y.z>. Each deploy/k8s/deploy.sh run tags the Docker images with both the explicit version (:${VERSION}) and :latest, so historical builds remain pullable for rollback.

packages/akb-mcp-client (the akb-mcp npm proxy) follows its own npm semver lifecycle and is not tied to the product version.

License

The AKB backend, frontend, and deployment manifests are licensed under the Business Source License 1.1 — source-available, with an Additional Use Grant that permits production use (commercial or non-commercial) up to a seat-count threshold, automatically converting to Apache License 2.0 four years after each version's first public release.

The npm akb-mcp proxy (packages/akb-mcp-client/) is separately licensed under the MIT License so it can be freely embedded in any agent client without restriction.

Free production use of the backend — you may deploy AKB in production, commercial or not, provided your aggregate deployment serves fewer than 100 Named Seats (distinct human user accounts in the users table, per deployment; service accounts and 90-day-inactive accounts excluded — see LICENSE for the precise definition).

Commercial license required for any of:

• Production use of the backend at or above 100 Named Seats.
• Offering AKB (modified or not) as a hosted service, on-premises product, embedded component, or rebranded distribution to third parties — regardless of seat count.

Trademarks — "AKB", "Dnotitia", and "Seahorse" are trademarks of Dnotitia, Inc. The software license does not grant trademark rights. Forks and derivative works must be distributed under a different name. See TRADEMARKS.md.

For commercial licensing, the rationale behind the BSL transition, or trademark permission requests, see LICENSE-CHANGE.md or contact [email protected].

Security

Found a vulnerability? See SECURITY.md — please report privately, not via public issues.

Contributing

See CONTRIBUTING.md.