feat(voice): configurable ElevenLabs max conversation duration by heavygee · Pull Request #230 · heavygee/hapi

heavygee · 2026-10-08T12:16:43Z

Fixes #229.

Auto-created ConvAI agents never set conversation.max_duration_seconds, so ElevenLabs applied its 600s default and cut conversations off.

New ELEVENLABS_MAX_DURATION_SECONDS, default 1800 (bounds the cost of a dead-open conversation; not the 7200 ceiling). Validated 60-7200; invalid values fail the token request with a clear message.
Applied at agent creation; reconciled onto existing hub-managed agents (found by name) once per process. Operator-supplied agents keep their own cap.
Tests assert the create payload carries the default and the configured value, plus PATCH reconciliation and rejection.
Verified live: created a temp agent via the payload (1800), drifted to 600, reconciled via the hub's PATCH body (1800, prompt/voice preserved), deleted it. API itself returns 400 on 7201.
Not done: friendlier agent names (needs a voice-name lookup).

🤖 Generated with Claude Code

Auto-created ConvAI agents never set conversation.max_duration_seconds, so ElevenLabs applied its 600s default and cut voice conversations off. Add ELEVENLABS_MAX_DURATION_SECONDS (default 1800, validated 60-7200). The cap is applied when the hub creates an agent and reconciled onto existing hub-managed agents (found by name) once per process. Operator-supplied agents (ELEVENLABS_AGENT_ID / voice map / customAgentId) keep their own cap. Invalid values fail the token request with a clear message instead of an API 422. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

chatgpt-codex-connector · 2026-10-08T12:18:24Z

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-12T17:58:33.171019Z	`89e3b5c`	New commits

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

Extend the hub-owner credentials panel (Settings -> Voice -> manage providers) with a numeric field for ELEVENLABS_MAX_DURATION_SECONDS, reusing the same env-wins-over-Settings overlay that already backs the ElevenLabs API key field. A Settings save takes effect on the very next token request in the same process; the env var still requires a restart. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

heavygee · 2026-10-12T17:44:41Z

Added the Settings UI on top of the env var, per operator request: Settings → Voice → manage providers (hub owner only) now has a "max conversation length" field under the ElevenLabs entry.

Backed by the same providerCredentials.ts overlay that already serves the ElevenLabs API key field: env wins and locks the field; otherwise a plain (unmasked — not a secret) value persisted to settings.json, held in the in-memory overlay, so a Settings save takes effect on the very next token request with no hub restart.
Client- and server-side range validation (60-7200), 400 on invalid, 409 if env-locked.
New hub tests cover default status, persist, reset-to-default, range rejection, env-lock conflict, and an end-to-end test proving a Settings PUT changes the very next token request's agent-creation payload in the same process.
Translated into en/zh-CN/ru.
Full suite: hub 1386 pass / 2 pre-existing unrelated failures (ambient FCM_SERVICE_ACCOUNT_PATH/Overseer env vars on this box affecting serverSettings.test.ts and titleSuggestion.test.ts, neither file touched by this change), shared 327 pass, web 3340 pass.

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: d23172d07c

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

…r process Codex review on #230 caught this: getOrCreateAgentIdForVoice's agentIdCache short-circuited before ever reaching ensureAgentOverrides, and ensureAgentOverrides's overridesEnsuredAgents gate was a one-time-forever marker that didn't account for the configured duration changing after the first reconciliation. Together, any voice already used once in the hub process would keep its old cap until a restart, contradicting the Settings UI's 'takes effect immediately' claim. Track the last-applied max duration per agent and compare it against the live setting on every call, including agentIdCache hits, so a Settings change reaches an already-cached agent on its very next token request. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

heavygee · 2026-10-12T17:55:39Z

Good catch from the Codex review — fixed in 89e3b5c.

getOrCreateAgentIdForVoice's agentIdCache returned a cached agent ID before ever reaching ensureAgentOverrides, and overridesEnsuredAgents was a one-time-forever marker that didn't account for the duration setting changing after the first reconciliation. Net effect: a voice already used once in the hub process would keep its old cap until a restart, which directly contradicted this PR's "takes effect on the next token request" claim.

Fix: track the last-applied max duration per agent (reconciledMaxDurationByAgent) and compare it against the live setting on every call, including cache hits. New regression test exercises the exact scenario: request a token for a voice, change the duration, request again for the same (now-cached) voice, and assert a PATCH with the new value fires without a re-create — then a third request confirms no further PATCH once it's current.

Dogfood the ElevenLabs max-duration fix + Settings UI before it merges. Thin soup tip resolves two trivial union conflicts against other layers in hub/src/config/settings.ts and hub/src/web/routes/voice.test.ts; no semantic changes vs the upstream-aimed feat/elevenlabs-max-duration tip. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

- Fix TS2769 typecheck failure: test assertions casting the credentials response narrowed the type to omit 'editable', which upstream CI's bun typecheck catches and this fork's (currently broken) CI did not. - An invalid ELEVENLABS_MAX_DURATION_SECONDS env value threw out of buildMaxDurationStatus(), which is shared by every provider's GET/PUT credentials status — breaking visibility into OpenAI/Deepgram/ Groq settings too, not just ElevenLabs. The token-issuance hot path (getConfiguredElevenLabsMaxDurationSeconds) still throws there by design; only the status/update endpoints now degrade gracefully. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

chatgpt-codex-connector · 2026-10-14T17:47:35Z

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, you can upgrade your account or add credits to your account and enable them for code reviews in your settings.

Nit from tiann#2017 follow-up review: the hint under the max-duration field rendered the literal '{min}-{max}' placeholders instead of '60-7200'. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

chatgpt-codex-connector · 2026-10-14T17:54:52Z

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, you can upgrade your account or add credits to your account and enable them for code reviews in your settings.

* feat(voice): configurable ElevenLabs max conversation duration Auto-created ConvAI agents never set conversation.max_duration_seconds, so ElevenLabs applied its 600s default and cut voice conversations off. Add ELEVENLABS_MAX_DURATION_SECONDS (default 1800, validated 60-7200). The cap is applied when the hub creates an agent and reconciled onto existing hub-managed agents (found by name) once per process. Operator-supplied agents (ELEVENLABS_AGENT_ID / voice map / customAgentId) keep their own cap. Invalid values fail the token request with a clear message instead of an API 422. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> * feat(voice): Settings UI for ElevenLabs max conversation duration Extend the hub-owner credentials panel (Settings -> Voice -> manage providers) with a numeric field for ELEVENLABS_MAX_DURATION_SECONDS, reusing the same env-wins-over-Settings overlay that already backs the ElevenLabs API key field. A Settings save takes effect on the very next token request in the same process; the env var still requires a restart. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> * fix(voice): reconcile max duration on cached agents, not just once per process Codex review on heavygee#230 caught this: getOrCreateAgentIdForVoice's agentIdCache short-circuited before ever reaching ensureAgentOverrides, and ensureAgentOverrides's overridesEnsuredAgents gate was a one-time-forever marker that didn't account for the configured duration changing after the first reconciliation. Together, any voice already used once in the hub process would keep its old cap until a restart, contradicting the Settings UI's 'takes effect immediately' claim. Track the last-applied max duration per agent and compare it against the live setting on every call, including agentIdCache hits, so a Settings change reaches an already-cached agent on its very next token request. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> * fix(voice): address #2017 review findings - Fix TS2769 typecheck failure: test assertions casting the credentials response narrowed the type to omit 'editable', which upstream CI's bun typecheck catches and this fork's (currently broken) CI did not. - An invalid ELEVENLABS_MAX_DURATION_SECONDS env value threw out of buildMaxDurationStatus(), which is shared by every provider's GET/PUT credentials status — breaking visibility into OpenAI/Deepgram/ Groq settings too, not just ElevenLabs. The token-issuance hot path (getConfiguredElevenLabsMaxDurationSeconds) still throws there by design; only the status/update endpoints now degrade gracefully. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> * fix(voice): interpolate duration range in the Settings hint Nit from #2017 follow-up review: the hint under the max-duration field rendered the literal '{min}-{max}' placeholders instead of '60-7200'. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> --------- Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>

heavygee · 2026-10-15T15:08:53Z

Merged upstream as tiann#2017 (squash 7b00df4). Closing the fork-stage copy; branch stays until Gate A cleanup deletes it.

chatgpt-codex-connector Bot reviewed Oct 12, 2026

View reviewed changes

Comment thread hub/src/web/routes/voice.ts

fix(voice): interpolate duration range in the Settings hint

a1ef87c

Nit from tiann#2017 follow-up review: the hint under the max-duration field rendered the literal '{min}-{max}' placeholders instead of '60-7200'. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

heavygee closed this Oct 15, 2026

heavygee deleted the feat/elevenlabs-max-duration branch October 15, 2026 15:09

heavygee mentioned this pull request Oct 15, 2026

CI: 'Test' GitHub Actions workflow has returned zero checks since ~2026-10-08 #277

Open

Navigation Menu

Repository navigation

feat(voice): configurable ElevenLabs max conversation duration - #230

feat(voice): configurable ElevenLabs max conversation duration#230
heavygee wants to merge 5 commits into
mainfrom
feat/elevenlabs-max-duration

heavygee commented Oct 8, 2026

Uh oh!

chatgpt-codex-connector Bot commented Oct 8, 2026 •

edited

Loading

Uh oh!

heavygee commented Oct 12, 2026

Uh oh!

chatgpt-codex-connector Bot left a comment

Uh oh!

Uh oh!

heavygee commented Oct 12, 2026

Uh oh!

chatgpt-codex-connector Bot commented Oct 14, 2026

Uh oh!

chatgpt-codex-connector Bot commented Oct 14, 2026

Uh oh!

heavygee commented Oct 15, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

heavygee commented Oct 8, 2026

Uh oh!

chatgpt-codex-connector Bot commented Oct 8, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codex Review Summary

Uh oh!

heavygee commented Oct 12, 2026

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

Uh oh!

heavygee commented Oct 12, 2026

Uh oh!

chatgpt-codex-connector Bot commented Oct 14, 2026

Uh oh!

chatgpt-codex-connector Bot commented Oct 14, 2026

Uh oh!

heavygee commented Oct 15, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

chatgpt-codex-connector Bot commented Oct 8, 2026 •

edited

Loading