Skip to content

fix(wiki): keep rebuilding after a page the model does not answer - #208

Merged
arthware-dev merged 2 commits into
feature/rc2from
fix/wiki-slow-model
Oct 13, 2026
Merged

arthware-dev merged 2 commits into
feature/rc2from
fix/wiki-slow-model

Conversation

@arthware-dev

Copy link
Copy Markdown
Contributor

Summary

  • A page whose model call times out no longer stops the wiki rebuild: that page stays as it was, and the run goes on with the rest.
  • Each page asks the model for at most 8,192 tokens, so a model caught in a loop can no longer hold the shared AI server for 15 to 30 minutes.

Test added 2 commits October 13, 2026 12:30
A page whose model call timed out raised `LLMTimeoutError`, which the
wiki did not catch: the run stopped there, and every page after it in
the selection or in the nightly sweep stayed stale until the next run.
Only an unreachable server was handled per page.

Each page now handles every model error the same way: it says which
page got no answer, leaves that page as it was, and the run goes on.
The wiki asked for each page without a token limit, so a model caught
in a repetition loop kept writing until the client timeout of 15
minutes, and the SDK's one retry could double that, all while holding
the AI server the rest of the household shares.

Each page call now asks for at most 8,192 tokens, about four times the
longest answer measured (the home page). A cut-off answer does not
parse, so the page stays as it was and the run goes on.
@arthware-dev
arthware-dev merged commit 1eafae9 into feature/rc2 Oct 13, 2026
1 check passed
@arthware-dev
arthware-dev deleted the fix/wiki-slow-model branch October 13, 2026 10:39
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant