Marc-André Bombeck f4c90dc134 fix(ai): pick the admin client from the configured base URL ...

The operator's global provider always built an OpenAI client, so an admin
who saved an Anthropic key and https://api.anthropic.com/v1 (a host the
settings allowlist accepts) had every AI call fail: an OpenAI-shaped body
posted at Anthropic's Messages API, with nothing in the error pointing at
the mismatch.

Route by hostname instead. api.anthropic.com and any subdomain of
anthropic.com get the same AnthropicClient the per-user ANTHROPIC arm
builds, with the same Claude model default; every other host keeps the
OpenAI client, which is what an OpenAI-compatible endpoint needs. The
match parses the URL rather than searching the string, so a lookalike
host cannot claim the Anthropic wire, and an unparseable value keeps the
OpenAI path it had before.

The chain tag stays admin-openai: it names the slot, not the wire, and it
is persisted in user chains and pinned by the OpenAPI enum. The wire is
already visible in the instance type the wide events read.

2026-09-02 16:17:10 +02:00

..

ai-budgets.test.ts

chore(release): v1.28.28 — OpenAI-compatible gateways work; a re-keyed night can no longer vanish

2026-07-11 10:36:02 +02:00

anthropic-client.test.ts

fix(ai): stop prefilling the assistant turn on Anthropic JSON calls

2026-09-02 10:28:05 +02:00

briefing-grounding.test.ts

feat(insights): weave ECG into the briefing narrative and daily rail

2026-07-17 08:19:17 +02:00

citation-coverage-logging.test.ts

fix(i18n): make the Korean locale actually work end to end

2026-08-26 20:00:44 +09:00

citation-enforcement.test.ts

Correct two test headers that described the removed cross-check

2026-07-19 05:53:05 +02:00

coach-prompt-v423.test.ts

feat(coach): quote the real BP range, address by name, render action cards

2026-06-27 14:40:07 +02:00

codex-client.test.ts

fix(coach): never 500 on a provider failure; fix Codex tool-call wire

2026-06-27 10:26:10 +02:00

codex-oauth.test.ts

fix(ai): complete Codex backend protocol implementation per spec

2026-05-09 14:36:12 +02:00

codex-slug-cache.test.ts

refactor(v1.4.15): apply simplify-review safe suggestions

2026-05-09 22:14:43 +02:00

codex-slug-fallback.test.ts

refactor(ai): take a structured message array instead of a string pair

2026-06-22 07:17:10 +02:00

codex-timeout.test.ts

fix(insights): stop the daily briefing going blank on slow generations

2026-06-27 12:37:25 +02:00

confidence-schema.test.ts

Delete the citation-enforcement chain that never ran

2026-07-19 05:50:57 +02:00

consent-guard.test.ts

feat(ai): add operator-shared central Codex connection with per-user opt-in

2026-07-10 00:10:56 +02:00

daily-briefing-schema.test.ts

feat(insights): verdict-first briefing, richer score cards, grounded opener rotation

2026-06-27 14:39:54 +02:00

effective-timeout.test.ts

fix(insights): honor the per-user AI timeout and keep the last good narrative

2026-06-28 12:42:35 +02:00

error-body-classification.test.ts

fix(security): scrub wide events at the builder, encrypt the replay cache

2026-07-19 01:28:20 +02:00

feedback-attribution.test.ts

feat(api): /api/insights/feedback persists thumbs with provider attribution

2026-05-10 01:30:06 +02:00

insight-generator-prompt.test.ts

Delete the citation-enforcement chain that never ran

2026-07-19 05:50:57 +02:00

insight-interpretation.test.ts

feat(insights): guideline interpretation registry for per-metric assessments

2026-07-07 00:36:25 +02:00

insight-response-schema.test.ts

Delete the citation-enforcement chain that never ran

2026-07-19 05:50:57 +02:00

json-extract.test.ts

fix(ai): stop prefilling the assistant turn on Anthropic JSON calls

2026-09-02 10:28:05 +02:00

legacy-payload-detection.test.ts

chore: voice + filename hygiene + wave-4 state wrap

2026-05-10 21:46:32 +02:00

local-client.test.ts

chore(release): v1.28.28 — OpenAI-compatible gateways work; a re-keyed night can no longer vanish

2026-07-11 10:36:02 +02:00

local-host-allowlist.test.ts

style: apply prettier across the tree

2026-06-19 22:20:31 +02:00

medical-reference-prompt.test.ts

feat(insights): verdict-first briefing, richer score cards, grounded opener rotation

2026-06-27 14:39:54 +02:00

medical-references.test.ts

style(prettier): sweep formatting across the tree

2026-05-10 11:05:40 +02:00

no-default-positivity-opener.test.ts

refactor(api,tz): one typed client for every call and one home for wall-clock math

2026-06-11 02:19:27 +02:00

openai-client.test.ts

fix: harden imports and daily medication state (#577)

2026-07-21 13:08:28 +02:00

openai-compatible-client.test.ts

Let the operator provider reach an allowlisted private host

2026-08-27 08:48:00 +02:00

openai-compatible-provider.test.ts

fix: preserve import job wipe ownership

2026-08-06 02:11:01 +02:00

provider-admin-host.test.ts

fix(ai): pick the admin client from the configured base URL

2026-09-02 16:17:10 +02:00

provider-chain.test.ts

feat(ai): add operator-shared central Codex connection with per-user opt-in

2026-07-10 00:10:56 +02:00

provider-contract.test.ts

test: route the global fetch stub onto the undici pinned engine

2026-06-28 15:18:20 +02:00

provider-health-ledger.test.ts

feat(ai): durable provider-health ledger and credential-expired surfacing

2026-06-04 00:50:13 +02:00

provider-runner.test.ts

Guard the operator's cap on fallback hops and reap dead summary promises

2026-08-14 05:53:17 +02:00

provider.test.ts

Surface effective AI availability on the ai-provider endpoint

2026-06-14 17:11:32 +02:00

rationale-prompt.test.ts

feat(ai): system prompt requires rationale per recommendation

2026-05-10 01:17:15 +02:00

rationale-schema.test.ts

Correct two test headers that described the removed cross-check

2026-07-19 05:53:05 +02:00

reference-id-schema.test.ts

feat(ai): schema requires rationale (window + comparedTo + deviation) per recommendation

2026-05-10 01:17:15 +02:00

resolve-chain.test.ts

style(prettier): sweep formatting across the tree

2026-05-10 11:05:40 +02:00

retry-correction.test.ts

fix(ai): align the corrective retry prompt with the validator that grades the reply

2026-07-19 11:01:57 +02:00

trend-annotations-schema.test.ts

feat(insights): verdict-first briefing, richer score cards, grounded opener rotation

2026-06-27 14:39:54 +02:00

vision-capability.test.ts

fix(ai): recognise Claude 5 and Fable models as vision-capable

2026-09-02 10:32:50 +02:00