ontoforge 3.9.0

Ontomesh — the ontology mesh for GraphRAG. Mine ontologies from logs, validate with SHACL, ship a hybrid retriever.

Ontomesh

The ontology mesh for GraphRAG. Mine ontologies from your logs, validate with SHACL, ship a hybrid retriever — without hand-crafting a single Turtle file.

Preview release. Public API surface is stabilising; expect occasional breaking changes until 1.0.

---

What's in the image

Image	ghcr.io/synaptixs/ontomesh:3.9.0 (and :latest)
Size	~316 MB compressed
Architectures	linux/amd64, linux/arm64 (Apple Silicon ready)
Base	python:3.12-slim
Runtime user	non-root ontomesh (uid 10001)
Process	single-process Flask via the ontoforge-wizard console script
Default port	5051 (configurable via ONTOMESH_PORT)
Persistent data	/data (mount a volume here)
Healthcheck	GET /health every 30 s

Run it

# Ephemeral run (resets on stop):
docker run --rm -p 5051:5051 ghcr.io/synaptixs/ontomesh:latest

# Persistent run (saved ontologies survive container restart):
docker volume create ontomesh-data
docker run -d --name ontomesh \
  -p 5051:5051 \
  -v ontomesh-data:/data \
  -e ONTOMESH_DATA_DIR=/data \
  ghcr.io/synaptixs/ontomesh:latest

Then open http://localhost:5051 in your browser.

Screenshots — a guided tour

A walkthrough from landing page to ontology-grounded reasoning search, in order.

1. Landing — the marketing entry point.

2. Model your domain — a guided, navy + lime-green Twilight wizard that walks you from a domain to a generated ontology in four phases — Define → Enrich → Build → After.

3. Run report — every run emits a branded HTML report: competency-question results, governance scorecard, semantic-loss analysis, and orphan-class checks.

4. Ask — pose a question — the read-only, ontology-grounded Ask console. Pick a flavor and provider (Ollama / OpenAI), set your sensitivity-tier ceiling, and optionally build an RDF subgraph for live SPARQL.

5. Ask — a cited, reasoned answer — the answer cites its sources, shows the exact read-only SQL that ran, materializes the result subgraph, and streams the full reasoning trace (plan → execute → relations → reason → synthesize).

6. Ask — live SPARQL over the result subgraph — interrogate exactly what grounded the answer with a SPARQL SELECT against the materialized RDF triples.

The wizard ships three themes — Twilight (default), Mono, and Light — switchable from the “All steps” drawer.

What you get

URL	What it serves
/	Marketing landing — value prop + install snippet
/wizard	The 11-step modeling wizard (Domain → Entities → Events → Relationships → Log Discovery → CQs → Rules → Generate → Evolution Review → Compliance → Vector Retrieval)
/projects	Dashboard of every ontology you've saved — one-click resume
/help	Per-phase help pages explaining each step
/health	JSON liveness probe
/api/events/stream	Server-Sent Events feed (drift + metric updates)
/api/ontologies	REST endpoint for the saved-ontologies store
/api/templates	REST endpoint listing available starter templates

Configuration (env vars)

Variable	Default	What it does
ONTOMESH_HOST	0.0.0.0	Bind address
ONTOMESH_PORT	5051	Bind port
ONTOMESH_DATA_DIR	/data	Where SQLite + saved ontologies + generated outputs live
ONTOMESH_DB_URL	(unset → SQLite at $ONTOMESH_DATA_DIR/ontologies.db)	Set to postgresql://user:pass@host:5432/dbname to use Postgres instead

What ships inside

Pipeline phases	8 build phases + 8 enrichment phases (log mining L4-L13, schema inference, multi-target generation)
Validation	SHACL 1.1 shapes + materialised inference with prov:wasDerivedFrom lineage
Reasoning	OWL-RL · SWRL · Datalog (Rulewerk)
Output formats	OWL/Turtle · SHACL · JSON-LD · SKOS · materialisation report
Wizard surfaces	10 starter industries (telecom, healthcare, finance, manufacturing, retail, energy-utilities, government, insurance, logistics, pharmaceuticals)
Discovery	Drain3 templates · per-service HMMs · Granger / transfer-entropy gating · PMI co-occurrence
Database backends	SQLite (default) · PostgreSQL · MySQL · MSSQL · Oracle · DB2
LLM adapters	Anthropic · OpenAI · Vertex AI · Ollama · OCI
Vector backends	FAISS · Chroma (scaffolded: Qdrant · pgvector)

First-time experience

The first time you load the wizard with an empty session, a welcome modal offers ten starter industries. Pick one to populate the session with a real ontology you can edit, or "Start with a blank session" to build from scratch.

Persistent SQLite vs. Postgres

The image defaults to SQLite-on-disk. For multi-host or multi-replica deployments, set ONTOMESH_DB_URL to a Postgres connection string — the wizard switches transparently. Both backends share the same wire schema.

Source · documentation · deployment recipes

The full developer documentation, deployment guides (Compose / Fly.io / Render / Cloud Run), and source code live in the repository:

https://github.com/synaptixs/ontomesh

Specifically:

• deploy/README.md — picking a deployment target
• docs/integrate.md — connecting to your own database
• compose.yml + deploy/Caddyfile — single-host TLS recipe
• fly.toml · render.yaml · deploy/cloudrun.yaml — managed-runtime configs

Feedback

File a tester-feedback issue — the form has a short three-option dropdown (Bug · Friction · Idea · Docs), optional repro/expected sections, and auto-applies the tester-feedback label. Or message the team directly.

Looking for what changed when? See CHANGELOG.md.

---

Ontomesh v3.9.0 · Apache-2.0 · Liquid Glass theme · Reasoning Search · gunicorn · Redis SSE · Prometheus · cosign-signed · OWL 2 · SHACL · PROV-O · SKOS · JSON-LD · 1000+ tests · Docker, Compose, Fly.io, Render, Cloud Run