# VelesDB Memory [Health: Active] **Category:** 🧠 Knowledge & Memory **Repository:** https://github.com/cyberlife-coder/VelesDB **GitHub Stars:** 91 **Views:** 0 **Installs:** 0 **Upvotes:** 0 **Directory Page:** https://allmcps.com/mcp/velesdb-memory ## Description Offline agentic memory: remember/recall/relate/forget/why over a fused vector+graph+columnar engine ## Claude Desktop Quick Installation Install path detected from listing signals. Uses `uvx` (confidence: high): ```json "mcpServers": { "velesdb-memory": { "command": "uvx", "args": ["velesdb"] } } ``` ## Documentation & README

VelesDB

VelesDB

One ~14 MB binary fuses vector + graph + columnar under a single query language β€” with an agent memory that shows its evidence and a deterministic context compiler that cuts your real, billed token spend.
Local-first: nothing leaves the machine, no LLM and no API key in the memory path. Every number below links to a committed harness you can rerun.

CI Crates.io PyPI npm Codacy coverage License
Quick start • Proof • Limitations • Architecture • Roadmap • velesdb.com

--- ## Start here β€” three commands that work ```bash pip install velesdb curl -O https://raw.githubusercontent.com/cyberlife-coder/VelesDB/main/examples/python/hello_velesdb.py python hello_velesdb.py ``` To search your own text instead of hand-written vectors, install the opt-in local adapter (`pip install "velesdb[embed-sentence-transformers]"`) and run [`hello_velesdb_text.py`](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/examples/python/hello_velesdb_text.py); its first run downloads `all-MiniLM-L6-v2`. Expected output, byte-for-byte ([read the script](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/examples/python/hello_velesdb.py) β€” no server, no embedding model): ``` Query: "tech" score=1.000 Rust 1.89 release notes score=0.600 AI-generated jazz: the new wave score=0.000 Best ramen in Tokyo Query: "tech + music" score=0.990 AI-generated jazz: the new wave score=0.707 Rust 1.89 release notes score=0.707 Miles Davis discography ``` An embedding model determines the vector dimension, while your similarity semantics determine the metric. Both are fixed when a collection is created; to change either, create a new collection and re-index your documents. **Give your agent a persistent memory β€” three more commands:** ```bash cargo install velesdb-memory # the local MCP memory server claude mcp add velesdb-memory -- ~/.cargo/bin/velesdb-memory # any MCP client works curl -L https://github.com/cyberlife-coder/VelesDB/releases/latest/download/velesdb-skills.tar.gz | tar -xz -C ~/.claude/skills/ ``` No Rust toolchain? `npm i @wiscale/velesdb-memory-node`, or grab a prebuilt `.mcpb` bundle from the [official MCP Registry](https://registry.modelcontextprotocol.io/?q=velesdb-memory) (`io.github.cyberlife-coder/velesdb-memory`).
Other paths β€” always-on hooks, shared daemon, Rust, Docker, WASM, REST **Memory used *continuously*, not just available:** [`integrations/agent-hooks/`](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/integrations/agent-hooks/README.md) wires five Claude Code hooks β€” `SessionStart`/`Stop`/`PreCompact` resume and save the working context, `PreToolUse` requires successful recall before an opted-in repository edit, and `PostToolUse` both records that recall and compiles an oversized tool result *before* it enters the transcript. One global install covers every project without enabling the edit guard outside explicitly configured repositories. **One memory shared by several clients** (Claude Code, Codex CLI, Claude Desktop, Windsurf, Devin CLI): [`scripts/install-memory-daemon.sh`](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-memory/README.md#http-transport-multi-client) runs `velesdb-memory` as a single local daemon β€” HTTPS by default, with a natively generated local CA. **Cargo (Rust + REST server):** `cargo install velesdb-server velesdb-cli` β€” **Docker** (multi-arch linux/amd64 + linux/arm64): `docker run -d -p 8080:8080 -v velesdb_data:/data --name velesdb ghcr.io/cyberlife-coder/velesdb:latest`, then `curl http://localhost:8080/health`. **Browser / edge:** the WASM build is ~710 KB gzipped and runs entirely client-side ([TypeScript SDK](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/sdks/typescript)). **REST:** 54 REST endpoints ([OpenAPI spec](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/openapi.yaml)). Full matrix: [installation guide](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/guides/INSTALLATION.md).
--- ## Why VelesDB - **One database instead of three.** Vectors for *"what feels similar"*, a graph for *"what is connected"*, typed columns for *"what I know for sure"* β€” normally three deployments, three query languages, and glue code. Here it is one binary and [one language](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/VELESQL_SPEC.md). - **A memory that can be audited, not just queried.** Every recall can show the evidence behind it; every compression decision carries a rule id, a reason, and a risk level. Deterministic by construction β€” no model in the write path, so no drift and nothing to re-litigate. - **Local-first is a sovereignty decision, not a latency one.** No cloud, no API key, no data processor: air-gapped if you want it, in your jurisdiction by default. [Why that matters](https://dev.to/wiscale-fr/i-built-a-database-in-france-because-the-cloud-act-makes-eu-data-sovereignty-impossible-5325) Β· [positioning in depth](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/WHY_VELESDB.md). ## How it works, in plain terms Four things happen, and none of them calls an AI provider. **1 Β· It stores facts, not conversations.** You give it one statement β€” *"the API port is 6333 because 3000 collided with the web UI"* β€” and it lands in a local file store. No model call, nothing sent anywhere. **2 Β· It finds them by meaning.** Asking *"which port did we settle on"* reaches that fact even though none of the words match. A local embedding model turns text into coordinates; close meaning means close coordinates. **3 Β· It connects them, and that is the part a search engine cannot do.** Each fact is linked to the topics it mentions. `why()` starts from the best match and then **walks those links**, so it returns the answer *plus the facts that explain it* β€” including ones sharing no vocabulary with your question. > The links have to exist. Store facts one by one and the graph stays flat, so > `why()` behaves like a search. Hand a paragraph to `remember_extracted` and it > splits it into facts and wires the links for you. **4 Β· It compresses what is too big, before you pay for it.** Give the compiler your accumulated context and a token budget; it returns a smaller version with **one recorded decision per fragment** β€” kept, abstracted, or dropped β€” and a handle to fetch any original back verbatim. Same input, same bytes out, every time. That is what the [82.5 % below](#proof--three-numbers-each-tied-to-its-harness) measures. --- ## What no one else combines ### 1 Β· Three engines, one query | Engine | What it does | |---|---| | **Vector** | Semantic similarity (HNSW + AVX2/NEON SIMD) | | **Graph** | Typed relationships, BFS/DFS, native `MATCH` clause ([patterns](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/guides/GRAPH_PATTERNS.md)) | | **ColumnStore** | Typed columnar metadata filtering, secondary indexes | One statement crosses all three β€” similarity, relations and typed filters, no glue code: ```sql MATCH (doc:Document)-[:AUTHORED_BY]->(author:Person) WHERE similarity(doc.embedding, $question) > 0.8 AND author.department = 'Engineering' RETURN author.name, doc.title ORDER BY similarity() DESC LIMIT 5 ``` ### 2 Β· A memory that shows its evidence β€” `why()` Most "agent memory" is vector recall: it finds text that *looks like* your query. VelesDB connects memories with typed links, so it can answer *why* something happened by walking the graph to context that shares **no words** with your question β€” across process restarts, offline, no API key: ```python from velesdb import MemoryService # pip install velesdb mem = MemoryService("./agent_memory") # a real on-disk store; survives restarts reason = mem.remember("Robert is recovering from knee surgery") mem.remember("Booked the aisle seat on Robert's flight", links=[(reason, "because")]) # A *new* process, weeks later, reopens the same store and asks why: mem.why("why the aisle seat on Robert's flight?") # walks booking β†’ reason β€” recall() can't ``` > `MemoryService` defaults to the offline `hash` embedder: deterministic, but > **lexical rather than semantic**, so unrelated wording can score `0.000`. > Opening it now says so once on stderr. For meaning-based recall, pass > `embedder="ollama"`; follow [Real semantic recall in 5 minutes](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-memory/README.md#real-semantic-recall-in-5-minutes). ![recall() finds the booking but misses the reason; why() reaches it through typed links, across a session restart](https://raw.githubusercontent.com/cyberlife-coder/VelesDB/HEAD/examples/agent_memory/why_across_sessions.gif) Memories are permanent by default; `forget(id)` deletes one, `ttl_seconds` gives a fact a durable expiry. Every `remember` auto-stamps its storage day, so recency-weighted recall works with zero setup. Same wedge in **Python**, **Node**, the [**MCP server**](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-memory), and in-memory in the [**TypeScript SDK**](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/sdks/typescript). Proof it is not a weak-embedder trick β€” four runnable demos in which `recall` stays blind to the reason **even under a real semantic embedder** (`ollama` / `all-minilm`), because the reason is connected by a decision rather than by surface similarity: [`why_across_sessions.py`](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/examples/agent_memory/why_across_sessions.py) (survives a process restart) Β· [`why_magic_constant.py`](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/examples/agent_memory/why_magic_constant.py) (a business reason sharing no words with the code) Β· [`memory_builds_its_own_graph.py`](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/examples/agent_memory/memory_builds_its_own_graph.py) (raw prose in, auto-wired graph out) Β· [`why_magic_constant.mjs`](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-node/examples/why_magic_constant.mjs) (Node). Benchmark position, including LoCoMo and why cross-lab scores are not fairly comparable: [BENCHMARK.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-memory/BENCHMARK.md) Β· [Agent Memory guide](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/guides/AGENT_MEMORY.md). ### 3 Β· A deterministic context compiler Agents burn most of their budget re-reading redundant context. `compile_context` / `compile_transcript` (MCP, or `ContextCompiler` in Rust) shrink it with **no LLM and no network**: - **Deterministic** β€” the same input always compiles to the same bytes, asserted twice per run in every committed benchmark. That also yields a [byte-stable cache prefix](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-memory/examples/context_savings/real_measures/cache_prefix.mjs) provider prompt-caching can actually hit. - **Auditable** β€” `explain_compilation` gives every kept or dropped fragment a stable rule id, a reason and a risk level. - **Reversible** β€” over-budget content becomes a recoverable `ctx://source/` handle; `retrieve_context_source` brings the original bytes back on demand. - **Bounded** β€” it compresses only what your agent explicitly hands it, never the harness's system prompt, and nothing enters recallable memory without an explicit `remember`. Code, URLs, numbers and negative constraints survive verbatim. The [`velesdb-context-optimizer` skill](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/skills/velesdb-context-optimizer/SKILL.md) teaches the workflow β€” including when *not* to compress. --- ## Proof β€” three numbers, each tied to its harness No figure here is an estimate from a slide; each links to the log or script in this repo that produced it. | Claim | Measured | Harness | |---|---|---| | Real **billed dollars** saved, same agent session sent raw vs compiled (real Claude billing, deterministic fact-checklist grader β€” no LLM judge) | **21.9 %** at real Retina screenshot weight, quality at parity (23.0/23 facts both arms) | [real-session-benchmark](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/examples/real-session-benchmark#billed-campaign-results-2026-07-19-cli-runner-claude-sonnet-5) Β· [raw logs](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/examples/real-session-benchmark/results/2026-07-19-vibe-cli/) | | Real (cl100k) **input-token savings** on a committed 12-turn agent-session corpus | **82.5 %**, compiled in ~0.5 ms mean stateless (~27 ms with source persistence on) | [context_savings](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-memory/examples/context_savings) | | **Vector search** latency on the full production path (VelesQL β†’ HNSW β†’ WAL ON β†’ payload hydration) | **450 us** p50 (10K/384D, recall β‰₯ 96 %) | [docs/BENCHMARKS.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/BENCHMARKS.md) | > Same campaign, less flattering: **10.9 %** on cropped screenshots, **14.7 %** on a 36-turn day-scale arc, **15.1 %** input tokens on the direct Messages API, and **2.5 %** for the no-screenshots variant β€” that spread *is* the measured value of the media mechanisms, so we publish it as prominently as the headline. [Honest reading, limitations and full protocol](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/examples/real-session-benchmark#honest-limitations). Every number on this page is CI-guarded by a [promise contract](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/reference/promise-contract.json) that pins the README to its committed sources.
The full measurement tables β€” billed A/B runs, retrieval quality, engine micro-benchmarks **Billed A/B sessions** (2026-07-19, claude-sonnet-5; raw logs committed verbatim): | Session | Runner | $ saved | Quality (raw vs compiled) | |---|---|---|---| | 19-turn feature session, cropped screenshots | Claude CLI | **10.9 %** | 22.8/23 vs 23.0/23 facts | | Same session, real Retina-weight screenshots | Claude CLI | **21.9 %** | 23.0/23 vs 23.0/23 | | 36-turn day-scale session | Claude CLI | **14.7 %** | 49.6/50 vs 49.2/50 * | | 19-turn session, direct Messages API | API | **15.1 %** input tokens | 23.0/23 vs 23.0/23 | > \* Two turns' grading key was later found defective (both arms scored full marks there; the parity conclusion stands) β€” [disclosure](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/examples/real-session-benchmark#billed-results-2026-07-19-all-real-executions). Over a [36-turn session](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/examples/real-session-benchmark#long-session-36-turns--context-window-headroom) compiled context grows **1.7Γ— slower**, so one session lasts far longer before hitting the window. **Memory retrieval quality**, public test sets, no AI grader in the loop: **+7.2 pts** multi-hop (HotpotQA), **+9.7 pts** time-scoped recall (TimeQA), **+29 pts** on a controlled task needing both engines at once β€” [BENCHMARK.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-memory/BENCHMARK.md). **End-to-end search** (canonical): search p50 **450 us** (10K, 384D, WAL ON) Β· SIMD dot product **21.7 ns** (768D, AVX2) Β· Recall@10 balanced 98.8 % Β· quantization PQ (8–32x), RaBitQ (32x), SQ8 (4x), Binary (32x) β€” [scope & caveats](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/guides/QUANTIZATION.md). **Index-only micro-benchmarks** (no WAL, no payload, hot cache β€” *not* comparable to the end-to-end figure above), each reproducible with `cargo bench -p velesdb-core --bench `: HNSW Search index-only (10K/768D, k=10) **55 us** (`hnsw_benchmark -- hnsw_search_latency`) Β· SIMD Dot Product (768D, AVX2) **21.7 ns** (`simd_benchmark`) Β· Recall@10 accurate mode **100%** (`recall_benchmark`) Β· BM25 Sparse Search index-only (10K docs, top-10) 57.6 us (`sparse_benchmark -- top10_10k_corpus`). | Search mode | ef_search | Recall@10 | Use case | |---|---|---|---| | Fast | 64 | 92.2% | Real-time suggestions, typeahead | | Balanced (default) | 128 | 98.8% | Production search, RAG pipelines | | Accurate | 512 | 100% | Evaluation, ground truth comparison | **Distance metrics** β€” 5 with SIMD acceleration (AVX-512, AVX2, NEON), at 768D/AVX2 on hot cache: Cosine 33 ns Β· Euclidean 20 ns Β· Dot Product 22 ns Β· Hamming 36 ns Β· Jaccard 35 ns. **ColumnStore** β€” typed columnar filtering, **130x faster** than JSON scanning at 100K rows on the i9-14900KF reference (`JSON scan 3.84 ms β†’ ColumnStore 29.5 us`). The ratio is hardware-dependent: on Apple Silicon (M5 Pro, 2026-07-20) the JSON scan itself runs ~2.8Γ— faster, so the same bench measures ~50–105x while the ColumnStore's absolute time holds (~27 Β΅s). > **Provenance:** Intel Core **i9-14900KF** (x86_64, AVX2). Per-machine figures vary; Apple-Silicon cross-checks, the SIFT1M standardized ANN run and the full methodology live in [docs/BENCHMARKS.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/BENCHMARKS.md). Reproduce the end-to-end figure with `python benchmarks/velesdb_benchmark.py --recall`.
--- ## Pick your entry point | I want to… | Use | Notes | |---|---|---| | Try it in one file | [`velesdb`](https://pypi.org/project/velesdb/) (Python 3.9+) | Fastest onboarding path | | Embed the engine | [`velesdb-core`](https://crates.io/crates/velesdb-core) (Rust) | The engine itself | | Give my agent memory | [`velesdb-memory`](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-memory) | MCP server + context compiler, any MCP client; `.mcpb` bundles on the [MCP Registry](https://registry.modelcontextprotocol.io/?q=velesdb-memory) | | Call it from Node | [`@wiscale/velesdb-memory-node`](https://www.npmjs.com/package/@wiscale/velesdb-memory-node) | Memory wedge ([full engine via server + TS SDK](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-node/README.md#need-the-full-engine)) | | Run it in a browser | [`@wiscale/velesdb-sdk`](https://www.npmjs.com/package/@wiscale/velesdb-sdk) | WASM, ~710 KB gzipped, fully client-side | | Serve it over HTTP | [`velesdb-server`](https://crates.io/crates/velesdb-server) | 54 REST endpoints β€” [API reference](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/reference/api-reference.md) Β· [OpenAPI](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/openapi.yaml) Β· [server security](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/guides/SERVER_SECURITY.md) | | Ship on mobile/desktop | [`velesdb-mobile`](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-mobile) Β· [Tauri plugin](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/tauri-plugin-velesdb) | iOS / Android / desktop | Tool parity per surface is published honestly β€” including where a surface is still behind: [memory crate README](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-memory/README.md). Worked examples: [examples/](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/examples/README.md).
Serving it over HTTP β€” the 54 REST endpoints, by category | Category | Key Endpoints | |----------|--------------| | **Collections** | `POST /collections`, `GET /collections`, `GET/DELETE /collections/{name}` | | **Points** | `/collections/{name}/points`, `/collections/{name}/points/raw`, `/collections/{name}/points/scroll`, `/collections/{name}/stream/insert`, `/collections/{name}/stream/enable`, `/collections/{name}/points/{id}/relations`, `/collections/{name}/points/{id}/ttl`, `/collections/{name}/relations` | | **Search** | `/collections/{name}/search`, `/collections/{name}/search/batch`, `/collections/{name}/search/hybrid`, `/collections/{name}/search/text`, `/collections/{name}/search/multi`, `/collections/{name}/search/ids`, `/collections/{name}/match` | | **Graph** | `/collections/{name}/graph/edges`, `/collections/{name}/graph/edges/{id}`, `/collections/{name}/graph/edges/count`, `/collections/{name}/graph/traverse`, `/collections/{name}/graph/traverse/stream`, `/collections/{name}/graph/traverse/parallel`, `/collections/{name}/graph/nodes`, `/collections/{name}/graph/nodes/{id}/degree`, `/collections/{name}/graph/nodes/{id}/edges`, `/collections/{name}/graph/nodes/{id}/payload`, `/collections/{name}/graph/search` | | **Indexes** | `GET/POST /collections/{name}/indexes`, `DELETE /collections/{name}/indexes/{label}/{property}`, `/collections/{name}/index/rebuild` | | **VelesQL** | `/query`, `/aggregate`, `/query/explain` | | **Admin** | `/health`, `/ready`, `/metrics`, `/guardrails`, `/collections/{name}/stats`, `/collections/{name}/config`, `/collections/{name}/flush`, `/collections/{name}/analyze`, `/collections/{name}/empty`, `/collections/{name}/sanity`, `/collections/{name}/compact`, `/collections/{name}/vacuum` | > **Full API reference:** [docs/reference/api-reference.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/reference/api-reference.md) | **OpenAPI spec:** [docs/openapi.yaml](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/openapi.yaml) | **Server security:** [docs/guides/SERVER_SECURITY.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/guides/SERVER_SECURITY.md)
## How it compares | | **VelesDB** | Chroma | Qdrant | pgvector | |---|---|---|---|---| | **Architecture** | Vector + graph + columnar, unified | Vector only | Vector + payload | Vector extension for PostgreSQL | | **Metadata filtering** | Typed ColumnStore + secondary indexes | JSON scan | JSON payload | SQL | | **Graph support** | Native (`MATCH` clause) | No | No | No | | **Query language** | VelesQL (SQL + NEAR + MATCH) | Python API | JSON API / gRPC | SQL + operators | | **Embeddings from text** | Opt-in [local / OpenAI adapters](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-python/python/velesdb/embed.py); no bundled model | [Embedding functions](https://docs.trychroma.com/docs/embeddings/embedding-functions), with a default local model in Python/TypeScript | Opt-in client-side [FastEmbed](https://qdrant.tech/documentation/fastembed/fastembed-semantic-search/) | None; bring vectors from an external model | | **Deployment** | Embedded / Server / WASM / Mobile | Server (Python) | Server (Rust) | Requires PostgreSQL | | **Binary size** | ~14 MB | ~500 MB (with deps) | ~50 MB | N/A (PG extension) | | **Browser / Mobile** | Yes / Yes | No | No | No | | **Offline / Local-first** | Yes | Partial | No | No | > **Sweet spot:** vector + graph + structured filtering in one engine, local-first, auditable. **Not the best fit (yet):** a managed cloud service with a multi-node distributed cluster. Competitor figures are typical public ranges, not a head-to-head run we performed β€” [run your own](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/BENCHMARKS.md). Detailed comparison against agent-memory products (Mem0, Zep, Letta), as of mid-2026: [docs/WHY_VELESDB.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/WHY_VELESDB.md). ## VelesDB Premium β€” the enterprise control plane The core engine is source-available and stays that way. **Premium** adds the company-grade layer on top of the same binary, for organizations running agent fleets on sensitive data: **RBAC** on every endpoint including the memory and context-compiler surfaces Β· **audit trail** (who, what, when β€” metadata only, GDPR-conscious) with forensic replay Β· **multi-tenancy** with hard per-tenant isolation and two-level deletion rights Β· **clustering and air-gapped deployment** Β· a **WebAdmin** UI for operators. Pricing on quote β€” **contact@wiscale.fr** Β· [velesdb.com](https://velesdb.com). Built by [Wiscale](https://wiscale.fr) (France; GDPR and data-sovereignty native). --- ## Known limitations β€” honest boundaries The items below are deliberate trade-offs or Premium-tracked features, **not correctness gaps** β€” the Community Edition is production-ready for single-node, local-first deployments. We publish them next to the strengths, including the ones we have not fixed yet. | # | Limitation | Scope | Tracked | |---|---|---|---| | 1 | **Single writer per collection** β€” WAL is serialized; concurrent writers contend on the same fsync lock. | Design trade-off (local-first, crash-safe by default). Read throughput is unaffected. | Concurrent WAL writer planned for [Premium](#velesdb-premium--the-enterprise-control-plane). See [docs/CONCURRENCY_MODEL.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/CONCURRENCY_MODEL.md). | | 2 | **No distributed replication** β€” single-node; no Raft, no sharding, no automatic failover in Core. | Deliberate: the sweet spot is local-first / embedded. | Raft-based replication tracked for Premium. | | 3 | **No advanced RBAC / multi-tenant isolation in Core** β€” Core ships the `DatabaseObserver` enforcement seam β€” live on every HTTP read path since 3.10.0 and still current in 4.0.0 β€” not the policy engine. | Core ships the hook, not the policy engine. | [Premium](#velesdb-premium--the-enterprise-control-plane) feature. | | 4 | **WASM MATCH limited to 2 hops** β€” 3+ hop `MATCH` works fully in native builds. | Browser-build scope limit, not a correctness issue. | Tracked. | | 5 | **SIFT1M fingerprint sidecar not yet committed** β€” the loader falls back to TOFU mode until the reference machine commits the pinned hashes. | Not a correctness issue β€” shape validation still applies. | Bootstrap shipped; sidecar pending. | | 6 | **No head-to-head Docker Compose benchmark vs Qdrant / Chroma / FAISS yet** β€” SIFT1M already gives literature-comparable numbers. | Side-by-side numbers need infrastructure not frozen yet. | Tracked. | | 7 | **Context-compiler tool parity varies by surface** β€” the MCP server and Rust have the full set; Node, Python and WASM are partially behind, and the WASM working contexts are intra-session only. | Binding scope, not an engine gap; MCP covers any client meanwhile. | Per-surface table in the [memory crate README](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/crates/velesdb-memory/README.md). | Internal technical limitations (query-planner approximations, plan-cache semantics): [docs/reference/KNOWN_LIMITATIONS.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/docs/reference/KNOWN_LIMITATIONS.md). ## Contributing & contact **Quality bar:** `cargo test --workspace` β€” 9k+ tests across Rust, TypeScript and Python run in CI on every merge; exact commands in [QUALITY_BAR.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/QUALITY_BAR.md). Contributions welcome β€” start with [CONTRIBUTING.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/CONTRIBUTING.md) and the [good first issues](https://github.com/cyberlife-coder/VelesDB/issues?q=is%3Aissue+is%3Aopen+label%3A%22good+first+issue%22). Security reports: [SECURITY.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/SECURITY.md). Roadmap: [ROADMAP.md](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/ROADMAP.md) Β· [Changelog](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/CHANGELOG.md) Β· [DeepWiki](https://deepwiki.com/cyberlife-coder/VelesDB). **License:** [VelesDB Core License 1.0](https://github.com/cyberlife-coder/VelesDB/blob/HEAD/LICENSE) (source-available). Premium: commercial license. **Contact:** contact@wiscale.fr Β· [velesdb.com](https://velesdb.com) The name nods to Veles, a deity of old Slavic myth β€” a keeper of hidden knowledge and boundaries.