feat: integrate WeKnora llmwiki knowledge flow
This commit is contained in:
1 parent
a0a75a9964
commit
e0c09f6704
63 files changed
+7045
-197
No files matched your search
@@ -257,7 +257,7 @@ Configuration priority:
|
||||
|
||||
### Gateway API (`app/gateway/`)
|
||||
|
||||
FastAPI application on port 8001 with health check at `GET /health`. Set `GATEWAY_ENABLE_DOCS=false` to disable `/docs`, `/redoc`, and `/openapi.json` in production (default: enabled).
|
||||
FastAPI application on port 8001 with health checks at `GET|HEAD /health`, `/api/health`, and `/api/langgraph/health`. These exact paths are always public in `auth_middleware.PUBLIC_HEALTH_PATHS`, including when authentication is enabled; reverse-proxy prefixes and trailing slashes are normalized before the whitelist check. Set `GATEWAY_ENABLE_DOCS=false` to disable `/docs`, `/redoc`, and `/openapi.json` in production (default: enabled). Regression coverage: `tests/test_health_auth_bypass.py`.
|
||||
|
||||
**Routers**:
|
||||
|
||||
@@ -600,7 +600,7 @@ Operators upgrading an existing DB run the whole file (`USE <db>;` then execute)
|
||||
### Tool System (`packages/harness/deerflow/tools/`)
|
||||
|
||||
`get_available_tools(groups, include_mcp, model_name, subagent_enabled)` assembles:
|
||||
1. **Config-defined tools** - Resolved from `config.yaml` via `resolve_variable()`
|
||||
1. **Config-defined tools** - Enabled entries are resolved from `config.yaml` via `resolve_variable()`; entries with `enabled: false` are omitted before import and are not registered with the agent
|
||||
2. **MCP tools** - From enabled MCP servers (lazy initialized, cached with mtime invalidation)
|
||||
3. **Built-in tools**:
|
||||
- `present_files` - Make output files visible to user (only `/mnt/user-data/outputs`)
|
||||
@@ -613,6 +613,7 @@ Operators upgrading an existing DB run the whole file (`USE <db>;` then execute)
|
||||
- `tavily/` - Web search (5 results default) and web fetch (4KB limit)
|
||||
- `jina_ai/` - Web fetch via Jina reader API with readability extraction
|
||||
- `firecrawl/` - Web scraping via Firecrawl API
|
||||
- `configurable_search/` - Config-driven `POST` JSON `web_search` replacement. It deep-copies the configured payload and overwrites only `query`, supports HTTP/HTTPS plus configurable `verify_ssl`, timeout and result limit, normalizes nullable `recUuid`/`content`/`title`/`url` fields, and rewrites a result URL through `result_url_template` only when `recUuid` is non-empty. The required response envelope is `{"results": [...]}`. Tests use `httpx.MockTransport` and never require the deployment endpoint.
|
||||
|
||||
**ACP agent tools**:
|
||||
- `invoke_acp_agent` - Invokes external ACP-compatible agents from `config.yaml`
|
||||
@@ -921,9 +922,9 @@ Both endpoints are whitelisted in `auth_middleware._PUBLIC_EXACT_PATHS`. The sha
|
||||
|
||||
Both can be modified at runtime via Gateway API endpoints or `DeerFlowClient` methods.
|
||||
|
||||
### WeKnora Retrieval Skill
|
||||
### Conditional WeKnora LLMWiki
|
||||
|
||||
`skills/custom/weknora-knowledge-search/` is a read-only integration for a WeKnora tenant. The Skill itself only directs the model to two MCP tools, `weknora_search` and `weknora_list_knowledge_bases`; the credential remains inside the stdio MCP subprocess (`scripts/weknora_retrieval_mcp.py`). Its retrieval core groups document/FAQ knowledge bases by `embedding_model_id` and calls WeKnora's multi-KB `hybrid-search` API once per group, preventing invalid cross-embedding score comparisons. It does not discard a KB merely because the legacy list API reports `chunk_count=0`; hybrid-search decides its actual readiness. Wiki pages are opt-in and use their dedicated Wiki search route. In explicit DeerFlow sharing mode, `WEKNORA_DEERFLOW_SHARE_TOKEN_FILE` selects WeKnora's narrow `/api/v1/deerflow/*` capability API: only owner-published document/FAQ knowledge bases are discoverable and searchable, while Wiki is skipped so the skill cannot fall back to a user bearer-token endpoint. Shared document hits include a credential-free source-preview URL. The frontend source card recognizes it and opens a context dialog; its call to authenticated Gateway route `/api/weknora/source-context` is proxied to WeKnora's token-gated `/api/v1/deerflow/knowledge-bases/:id/chunks/:chunk_id/context`, returning only the matching chunk and two neighbours on either side. Never expose the share token to the browser or add it to a URL. Configure through the bundled `templates/extensions_config.fragment.json`; otherwise prefer `WEKNORA_API_KEY_FILE` for scoped-key deployments and `WEKNORA_BEARER_TOKEN_FILE` for legacy deployments. `WEKNORA_ALLOWED_KB_IDS` remains an optional allow-list. Tests: `tests/test_weknora_knowledge_search_skill.py`, `tests/test_weknora_source_preview.py`.
|
||||
`llmwiki.weknora.api_base_url` is the provider switch: non-empty enables WeKnora, empty preserves the legacy LLMWiki UI. WeKnora is independently deployed and owns PostgreSQL/Redis/object storage/vector storage/parsers/models; DeerFlow config contains only `api_base_url` and optional `web_base_url`. DeerFlow deliberately sends no WeKnora credential, so the WeKnora endpoint must be restricted to a trusted network and allow DeerFlow traffic itself or through its reverse proxy. There is intentionally no `default_embedding_model_id`: `WeKnoraClient` queries `/api/v1/models` and uses WeKnora's active Embedding model when creating a KB; Wiki creation additionally resolves an active KnowledgeQA model inside WeKnora for synthesis. The native BFF is `app/gateway/routers/llmwiki.py`; ownership/publication mappings live in `deerflow.persistence.llmwiki`; the browser only sees DeerFlow mapping IDs. Knowledge-base owners publish immediately (`private` -> `published`) without a review queue, and can unpublish. Knowledge cards navigate to a dedicated WeKnora-style detail route. Wiki-enabled bases render and default to `Wiki -> documents -> graph`; standard bases omit Wiki and render `documents -> graph`. The detail page supports file, URL and published manual-Markdown imports, browses real Wiki pages/content, proxies authenticated original-file previews, pages through `/chunks/{knowledge_id}`, and renders the Wiki page graph from `/knowledgebase/{kb_id}/wiki/graph`; document details use a right sheet rather than the removed KB modal. The credential-free proxy uses one WeKnora service identity, so per-user separation is enforced by DeerFlow mappings while all remote KBs physically share that identity's WeKnora workspace. `llmwiki_search` is conditionally registered in `tools.py`, is read-only, and rechecks the current DeerFlow user, agent `llmwiki_knowledge_base_ids`, and thread-context selection on every call. Agent bindings use an inline WeKnora-style searchable/grouped multi-select in the standard agent create/detail UI; new chats use the same interaction pattern as a compact composer dropdown, and selections are written to thread metadata/context and restored on reopen. The DeerFlow frontend is native React and the WeKnora frontend is not modified or embedded. Full deployment and behavior notes: `docs/LLMWIKI_WEKNORA_INTEGRATION_ZH.md`. Tests: `tests/test_llmwiki_weknora.py`.
|
||||
|
||||
### Embedded Client (`packages/harness/deerflow/client.py`)
|
||||
|
||||
@@ -1133,7 +1134,7 @@ not delete a built-in id because historical chain nodes persist those ids;
|
||||
custom role ids may be added and removed.
|
||||
|
||||
See `docs/` directory for detailed documentation:
|
||||
- [../README.md](../README.md) - Offline Docker wheelhouse preparation and startup workflow. The pinned versions in `../offline-runtime-requirements.txt` and `../start_backend.sh` must stay synchronized. `start_backend.sh` stays aligned with the deployment script and adds only the Redis wheel fallback line. The download target defaults to CPython 3.12 on Linux x86_64 and verifies dependency closure using wheelhouse only.
|
||||
- [../README.md](../README.md) - Offline Docker wheelhouse preparation and startup workflow. The pinned versions in `../offline-runtime-requirements.txt` and `../start_backend.sh` must stay synchronized. Startup checks `croniter`, `hindsight-client`, `aiomysql`, and Redis, installing missing packages from wheelhouse only; `aiomysql==0.3.2` includes the pinned `PyMySQL==1.2.0` transitive dependency. The download target defaults to CPython 3.12 on Linux x86_64 and verifies dependency closure using wheelhouse only.
|
||||
- [CONFIGURATION.md](docs/CONFIGURATION.md) - Configuration options
|
||||
- [ARCHITECTURE.md](docs/ARCHITECTURE.md) - Architecture details
|
||||
- [API.md](docs/API.md) - API reference
|
||||
|
||||
Reference in new issue
Block a user