UIengF opened a new pull request, #382:
URL: https://github.com/apache/hugegraph-ai/pull/382

   MCP refresh can use a different AI-service graph from the MCP target, while 
the toolset names and configuration guidance blur released and source behavior. 
In addition, login parsing differs between authentication and the Thin API 
write gate, and UI configuration reload can change the target during a refresh.
   
   This PR:
   
   - Names source toolsets `core` and `legacy`, retaining warning aliases for 
`v2_core` and `v1`. Toolset registration and metadata use one startup snapshot; 
invalid values prevent startup.
   - Deprecates the unused `HUGEGRAPH_AI_GRAPH_URL` override. Refresh probes 
the AI target-validation capability, sends the expected graph and graphspace, 
and rejects mismatches before execution. Failed probes never dispatch refresh; 
authentication and transient errors retain their classification and 
retryability.
   - Uses one login flag parser and the existing runtime lock for UI 
configuration reload, keeping refresh validation and execution on the same 
configured target.
   - Sends Gremlin example embeddings sequentially in batches of at most 10, 
including the LiteLLM path, and saves the index only after every batch 
succeeds. OpenAI-compatible embedding defaults use 10 for both sync and async 
calls.
   - Aligns bilingual READMEs, integration/release guidance and skills with 
implemented capabilities, and adds a client-owned interactive STDIO example. 
The documentation distinguishes published `1.7.1` from current source behavior.
   
   ### Validation
   
   - MCP suite: **861 passed, 17 skipped**; real STDIO subprocess smoke covers 
initialization, tool listing, inspection and structured queries against a local 
HTTP fixture.
   - Thin API, RAG API, configuration and example-index regressions: **95 
passed** on Python 3.10 in an isolated container using current source. New 
regressions fail before the fixes and pass after them: concurrent authenticated 
refresh/UI reload, LiteLLM batch limits/order, and failure in a later batch 
without saving an index.
   - Repository Ruff lint and format checks, and `git diff --check`: passed.
   - Earlier integration validation for the unchanged MCP contract included 
**53 live STDIO assertions**, **11 AI HTTP cases**, normal/repeated refresh, 
target/authentication/feature gates, and pinned-package/source-install checks.
   
   The new concurrency and provider-limit regressions use a controlled 
scheduler/provider boundary. The earlier live AI acceptance used isolated 
indexes and containers, with lifespan disabled to test HTTP gates 
independently; it does not validate startup background refresh gating. 
Database-instance identity remains a deployment responsibility when graph and 
graphspace names match. This PR does not publish packages or deploy services.
   


-- 
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]

Reply via email to