pediatric-ai-scribe-v3/docs
Daniel efa84ac0e6
Some checks failed
Forgejo Android APK / Root app tests (push) Successful in 47s
Forgejo Docker Build / Root app tests (push) Successful in 47s
Forgejo Android APK / Build signed APK (push) Successful in 2m3s
Forgejo Docker Build / Build Docker image (push) Successful in 10s
Forgejo Docker Build / Deploy to the host (push) Failing after 0s
docs: one knob per feature, and record how many excerpts each gets
Answering a fair question: needing two settings in two repositories to agree
before one number changes is not a design anyone chose. It is two codebases each
assuming it owned the decision, and the symptom was a caller asking for 30
excerpts and silently receiving 12.

rerank_results computes min(reranker_top_k or limit, limit), so RERANKER_TOP_K=0
reads as "however many the caller asked for". The app setting is now the only
knob. Verified: Learning asks 30 and receives 30, the assistant asks 8 and
receives 8.

Zero costs nothing extra — the reranker is billed on documents sent, which is
candidate_limit and unchanged; top_n only decides how many come back. A real
number there is now what it should always have been: an optional hard ceiling
for when someone deliberately wants one, not an invisible default.

docs/retrieval-tuning.md covers the per-feature budgets, why the assistant's are
so much smaller than Learning's, that My Resources deliberately shares the
Learning budget, how to read what actually happened from the MCP logs and the
grounding field, and why raising these is not free.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Dv6sqaY6Vq3ChZHMem3cnU
2026-09-11 15:28:01 +02:00
..
logic simplify speech and embeddings through litellm 2026-05-09 05:09:02 +02:00
ai-providers.md docs refresh current workflows 2026-05-09 00:40:45 +02:00
api-reference.md docs: recording, 24-hour retention, and the STT picker 2026-09-10 16:44:17 +02:00
ARCHITECTURE.md document architecture and harden rendering 2026-05-10 01:07:56 +02:00
architecture.md ci: route Android APK through Forgejo releases 2026-05-11 04:21:15 +02:00
authentication.md Fix Turnstile in the mobile app; drop it from login 2026-07-30 17:22:03 +02:00
CLINICAL_ASSISTANT.md feat: My Resources has a menu, a library and three downloads 2026-09-11 15:00:19 +02:00
configuration.md feat: My Resources has a menu, a library and three downloads 2026-09-11 15:00:19 +02:00
database.md docs align memory and audio backup behavior 2026-05-08 23:32:39 +02:00
deployment.md feat: ship reviewed prompt history, conversation limits and account protections 2026-09-07 04:01:01 +02:00
developer-guide.md feat: ship reviewed prompt history, conversation limits and account protections 2026-09-07 04:01:01 +02:00
DEVELOPMENT.md feat: ship reviewed prompt history, conversation limits and account protections 2026-09-07 04:01:01 +02:00
embeddings-setup.md simplify speech and embeddings through litellm 2026-05-09 05:09:02 +02:00
features-explained.md simplify speech and embeddings through litellm 2026-05-09 05:09:02 +02:00
global-prompt-administration.md fix: reviewer findings — mhchem macro wrapping, regenerate history dedupe, strict safe image allowlist, stale docs 2026-09-08 19:25:31 +02:00
improvements.md docs refresh current workflows 2026-05-09 00:40:45 +02:00
learning-hub.md feat: My Resources has a menu, a library and three downloads 2026-09-11 15:00:19 +02:00
migrations.md docs: rewrite architecture, authentication, configuration, deployment, ai-providers, speech, database, learning-hub, migrations, developer-guide for public audience 2026-04-15 00:26:38 +02:00
mobile-build.md ci: route Android APK through Forgejo releases 2026-05-11 04:21:15 +02:00
MODULE_CONVENTIONS.md document architecture and harden rendering 2026-05-10 01:07:56 +02:00
openid-setup.md feat: ED multi-stage UX, extensions polish, docs viewer + application-logic docs 2026-04-28 03:09:38 +02:00
ops-docs-ped-ai-and-milvus.md docs: Ops Docs page on how Ped-AI and the two Milvus servers work 2026-09-10 18:37:41 +02:00
retrieval-tuning.md docs: one knob per feature, and record how many excerpts each gets 2026-09-11 15:28:01 +02:00
SCALING.md document architecture and harden rendering 2026-05-10 01:07:56 +02:00
speech.md refactor: one place decides which bucket, on which S3, with which credentials 2026-09-10 17:09:13 +02:00
transcription-options.md docs: recording, 24-hour retention, and the STT picker 2026-09-10 16:44:17 +02:00