Semantica
Builds a knowledge graph from your documents without calling an LLM, then answers agents through graph traversal, causal chains, and decision records that keep their provenance.
What it provides
Connect an existing HTTPS deployment with an optional scoped token. Heavy jobs stay on that service instead of using bot memory.
Connection requirements
Install `semantica[explorer]` and run `python -m semantica.server` outside the bot, which listens on port 8000, then publish it through an authenticated HTTPS reverse proxy. Upstream ships no authentication at all and binds 0.0.0.0, so that proxy is the only thing checking callers: keep port 8000 itself unreachable from the internet by binding it to loopback or a private network and firewalling it, because anyone who can reach the host directly walks straight past the proxy and can read your knowledge graph and write new facts into it. SEMANTICA_SECRET_KEY does not add API authentication and is no substitute for the proxy. Keep Bearer selected and paste the proxy token below. The Explorer API routes, including the three below, are mounted only when the explorer imports succeed: if they fail the server still starts and answers 404 there, so confirm with `get /api/graph/stats` before relying on it. Point the graph and vector stores at persistent backends on that host — extraction, embeddings, and the graph itself stay there rather than in the bot.
- Protocol
- REST / OpenAPI
- Endpoint
https://semantica.example.com- Authentication
- bearer
post /api/enrich/extract {"text":"Apple CEO Tim Cook announced record earnings."}get /api/graph/node/acme_corp/neighbors?depth=2get /api/graph/statsWhere to add it
Open Dashboard Tools, select a running OpenClaw or Hermes instance, then use the card’s Install or Connect action. External services may require an HTTPS endpoint and credentials; the dashboard shows those fields before anything is saved.