bitflicker64 commented on code in PR #543: URL: https://github.com/apache/hugegraph-doc/pull/543#discussion_r4325061125
########## content/en/docs/quickstart/hugegraph-ai/ontogeny.md: ########## @@ -0,0 +1,91 @@ +--- +title: "Ontogeny" +linkTitle: "Ontogeny" +weight: 7 +--- + +`ontogeny` — Greek *onto-* (being, the root of *ontology*) + *genesis* (birth, development): in biology the word names the **full-life-cycle development of an individual organism**. The name is the thesis: an ontology here is not a static schema but a living thing — declared in Git, compiled into tables and a graph, and grown through use. It is a **self-improving operational ontology platform**: it models an enterprise's business semantics (objects, properties, links) and business actions (actions, functions, policies) as declarative, Git-native YAML, and serves them to applications, analytics and AI agents through one governed API — optionally projected onto [Apache HugeGraph](https://github.com/apache/hugegraph) for deep-graph queries. + +The whole system on one board — this is the console's Overview dashboard, and the map for everything below: + + + +*Top: the eight first-class DSL citizens (Action = the only write gate, EvalSuite = the selection box). Middle: the online pipeline (build → explore → act → control → audit) and the offline RSI loop, joined by the agent spine — its reads/writes cross the same gates, its traces and approvals flow back as evolution signals. Bottom: the supporting layer and the three cross-cutting invariants (ownership, one gate, Git as source).* + +## Key Features + +- **Git-native ontology** — modeling is YAML + pull requests; consensus happens in code review, not a locked GUI +- **Single-process kernel** — `ontogeny serve` = Registry + query engine + Action runtime on SQLite / Postgres +- **Agent-first** — a governed MCP tool surface at `/mcp`; agents and humans share one policy surface, writes pass human approval gates +- **Self-improving (RSI)** — observe → mutate (LLM-generated DSL diffs) → select (deterministic eval suites) → release (T0 auto-merge / T1 PR / T2 canary); *mutation is outsourced to the LLM, the selection function is not* +- **Constitutional guardrails** — machine proposals travel the same Git/CI channel as humans; policy relaxation, markings and tier definitions always require human review +- **Bilingual enterprise console** — React 19 + TypeScript: ontology canvas (browse *and* edit), graph exploration, action runner, audit, self-evolution console + +Modeling is code review, not form-filling — the ontology canvas shows every object, link, action and function of the domain, with references highlighted on hover: + + + +## Quick Start + +### Option 1: Docker (Recommended) + +```bash +git clone -b ontogeny https://github.com/apache/hugegraph-ai.git +cd hugegraph-ai + +docker compose up -d # self-contained demo (SQLite + seeded data) +# docker compose --profile postgres up -d # Postgres-backed +# docker compose --profile graph up -d # + HugeGraph projection +``` + +Console + API: **http://localhost:8000** — the first boot creates an `admin` account and prints the generated password to the container log (or preset `ONTOGENY_ADMIN_PASSWORD`). One command gets you a live dashboard seeded with a real manufacturing domain — the ontology model, the agent surface and the RSI loop at a glance: Review Comment: Important: The first boot does not generate a password. `ensure_admin()` in `server/ontogeny/auth/service.py` on the `ontogeny` branch uses `os.environ.get("ONTOGENY_ADMIN_PASSWORD") or "ontogeny@2026"`. The log line it prints says "default password ontogeny@2026". So every instance started with this command has the same published admin password, and compose publishes port 8000 on all interfaces. A reader who believes the password is random has no reason to change it. The CN page has the same sentence at `content/cn/docs/quickstart/hugegraph-ai/ontogeny.md` line 41 (“生成的初始密码”). Requested change: name the fixed default and say to override or change it before the port is reachable. For example: ```suggestion Console + API: **http://localhost:8000** — the first boot creates an `admin` account with the fixed default password `ontogeny@2026` (also printed to the container log). Set `ONTOGENY_ADMIN_PASSWORD` before the first start, or change the password before exposing port 8000. One command gets you a live dashboard seeded with a real manufacturing domain — the ontology model, the agent surface and the RSI loop at a glance: ``` CN line 41 can read: “首次启动会自动创建 `admin` 账号，初始密码为固定默认值 `ontogeny@2026`（容器日志中也会打印）。请在首次启动前设置 `ONTOGENY_ADMIN_PASSWORD`，或在对外暴露 8000 端口前修改密码。” ########## content/en/docs/quickstart/hugegraph-ai/ontogeny.md: ########## @@ -0,0 +1,91 @@ +--- +title: "Ontogeny" +linkTitle: "Ontogeny" +weight: 7 +--- + +`ontogeny` — Greek *onto-* (being, the root of *ontology*) + *genesis* (birth, development): in biology the word names the **full-life-cycle development of an individual organism**. The name is the thesis: an ontology here is not a static schema but a living thing — declared in Git, compiled into tables and a graph, and grown through use. It is a **self-improving operational ontology platform**: it models an enterprise's business semantics (objects, properties, links) and business actions (actions, functions, policies) as declarative, Git-native YAML, and serves them to applications, analytics and AI agents through one governed API — optionally projected onto [Apache HugeGraph](https://github.com/apache/hugegraph) for deep-graph queries. + +The whole system on one board — this is the console's Overview dashboard, and the map for everything below: + + + +*Top: the eight first-class DSL citizens (Action = the only write gate, EvalSuite = the selection box). Middle: the online pipeline (build → explore → act → control → audit) and the offline RSI loop, joined by the agent spine — its reads/writes cross the same gates, its traces and approvals flow back as evolution signals. Bottom: the supporting layer and the three cross-cutting invariants (ownership, one gate, Git as source).* + +## Key Features + +- **Git-native ontology** — modeling is YAML + pull requests; consensus happens in code review, not a locked GUI +- **Single-process kernel** — `ontogeny serve` = Registry + query engine + Action runtime on SQLite / Postgres +- **Agent-first** — a governed MCP tool surface at `/mcp`; agents and humans share one policy surface, writes pass human approval gates +- **Self-improving (RSI)** — observe → mutate (LLM-generated DSL diffs) → select (deterministic eval suites) → release (T0 auto-merge / T1 PR / T2 canary); *mutation is outsourced to the LLM, the selection function is not* +- **Constitutional guardrails** — machine proposals travel the same Git/CI channel as humans; policy relaxation, markings and tier definitions always require human review +- **Bilingual enterprise console** — React 19 + TypeScript: ontology canvas (browse *and* edit), graph exploration, action runner, audit, self-evolution console + +Modeling is code review, not form-filling — the ontology canvas shows every object, link, action and function of the domain, with references highlighted on hover: + + + +## Quick Start + +### Option 1: Docker (Recommended) + +```bash +git clone -b ontogeny https://github.com/apache/hugegraph-ai.git +cd hugegraph-ai + +docker compose up -d # self-contained demo (SQLite + seeded data) +# docker compose --profile postgres up -d # Postgres-backed Review Comment: Important: `docker compose --profile postgres up -d` starts a Postgres container, but Ontogeny keeps using SQLite. The image's `CMD` is `ontogeny serve --demo --data-dir /data ...` with no `--db`. Under `--demo`, `serve()` copies `DemoPaths.as_env()` into `os.environ`, and that sets `ONTOGENY_DB_DSN` to `sqlite+aiosqlite:///<data-dir>/ontogeny.db` (`prepare_demo` in `server/ontogeny/demo/bootstrap.py`). This happens even when the user exports a Postgres DSN for compose. The image also installs only `'.[mcp]'` (`deploy/Dockerfile`), not the `postgres` extra with `asyncpg`. A reader who picks the "Postgres-backed" line ends up with data in the SQLite volume `ontogeny-data`, not in Postgres. The graph line below has a similar gap: compose passes no `HUGEGRAPH_URL` to the `ontogeny` service, so `storage_provider` stays `sqlite` until a storage URL such as `http://hugegraph:8080` is set through `PUT /api/v1/admin/runtime-config`. The CN page has the same comments at lines 37-38 (“换 用 Postgres”, “追加 HugeGraph 图投影”). Requested change: drop the Postgres line until the image and the demo entrypoint support it, or describe what it actually does. For the graph line, add the step that connects the projection. For example: ```suggestion # docker compose --profile postgres up -d # starts a Postgres container only; the shipped image still runs the SQLite demo ``` ########## content/en/docs/quickstart/hugegraph-ai/ontogeny.md: ########## @@ -0,0 +1,91 @@ +--- +title: "Ontogeny" +linkTitle: "Ontogeny" +weight: 7 +--- + +`ontogeny` — Greek *onto-* (being, the root of *ontology*) + *genesis* (birth, development): in biology the word names the **full-life-cycle development of an individual organism**. The name is the thesis: an ontology here is not a static schema but a living thing — declared in Git, compiled into tables and a graph, and grown through use. It is a **self-improving operational ontology platform**: it models an enterprise's business semantics (objects, properties, links) and business actions (actions, functions, policies) as declarative, Git-native YAML, and serves them to applications, analytics and AI agents through one governed API — optionally projected onto [Apache HugeGraph](https://github.com/apache/hugegraph) for deep-graph queries. + +The whole system on one board — this is the console's Overview dashboard, and the map for everything below: + + + +*Top: the eight first-class DSL citizens (Action = the only write gate, EvalSuite = the selection box). Middle: the online pipeline (build → explore → act → control → audit) and the offline RSI loop, joined by the agent spine — its reads/writes cross the same gates, its traces and approvals flow back as evolution signals. Bottom: the supporting layer and the three cross-cutting invariants (ownership, one gate, Git as source).* + +## Key Features + +- **Git-native ontology** — modeling is YAML + pull requests; consensus happens in code review, not a locked GUI +- **Single-process kernel** — `ontogeny serve` = Registry + query engine + Action runtime on SQLite / Postgres +- **Agent-first** — a governed MCP tool surface at `/mcp`; agents and humans share one policy surface, writes pass human approval gates +- **Self-improving (RSI)** — observe → mutate (LLM-generated DSL diffs) → select (deterministic eval suites) → release (T0 auto-merge / T1 PR / T2 canary); *mutation is outsourced to the LLM, the selection function is not* +- **Constitutional guardrails** — machine proposals travel the same Git/CI channel as humans; policy relaxation, markings and tier definitions always require human review +- **Bilingual enterprise console** — React 19 + TypeScript: ontology canvas (browse *and* edit), graph exploration, action runner, audit, self-evolution console + +Modeling is code review, not form-filling — the ontology canvas shows every object, link, action and function of the domain, with references highlighted on hover: + + + +## Quick Start + +### Option 1: Docker (Recommended) + +```bash +git clone -b ontogeny https://github.com/apache/hugegraph-ai.git +cd hugegraph-ai + +docker compose up -d # self-contained demo (SQLite + seeded data) +# docker compose --profile postgres up -d # Postgres-backed +# docker compose --profile graph up -d # + HugeGraph projection +``` + +Console + API: **http://localhost:8000** — the first boot creates an `admin` account and prints the generated password to the container log (or preset `ONTOGENY_ADMIN_PASSWORD`). One command gets you a live dashboard seeded with a real manufacturing domain — the ontology model, the agent surface and the RSI loop at a glance: + + + +### Option 2: From Source + +```bash +# backend (with the MCP extra so /mcp is served) Review Comment: Minor: The EN "From Source" block starts with `uv venv` and never clones the `ontogeny` branch. Option 2 is an alternative to Option 1, so a reader who goes straight to it has no checkout and no `pyproject.toml` or `web/` to work with. The CN page has the clone and `cd` lines at `content/cn/docs/quickstart/hugegraph-ai/ontogeny.md` lines 48-49, so the two languages also differ. Requested change: add the same two lines as the CN page: ```suggestion git clone -b ontogeny https://github.com/apache/hugegraph-ai.git cd hugegraph-ai # backend (with the MCP extra so /mcp is served) ``` -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
