feat(docs): grounded NLQA + support surface + auto-generated docs index - #273
Merged
Conversation
Docs are auto-generated from the source specs (build_docs_index.py → docs-index.json, 43 docs) — a doc is a witness of source. The Docs & Support surface answers ONLY from that grounded corpus, cites every source, and abstains to support when a question isn't covered (fail-closed grounding — no answer without a citation). CI drift-guard (docs-index-fresh) regenerates + fails if the committed index drifted from source.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Links documentation ↔ NLQA ↔ support in a grounded, self-updating way (the capstone of the surfaces work).
tools/build_docs_index.py— auto-generates the grounded corpus (docs/surfaces/data/docs-index.json, 43 docs) from the source specs. A doc is a witness of source.docs/surfaces/docs-support.html— the grounded NLQA + support surface. Answers come only from the index, extractive + cited (no hallucination); grounding is graded GROUNDED / PARTIAL / ABSTAINED, and an ungrounded question abstains — 'I won't answer what I can't cite' — opening support and logging the doc gap instead of guessing. Routes lampstand → sherlock (IR) → InferenceGateway; receipted..github/workflows/docs-index.yml— drift-guard: regenerates the index and fails if the committed one drifted from source, so 'auto-generated' is a guarantee, not a hope.Try 'what is the capital of France?' — it abstains, because it isn't in the docs. That refusal is the feature.