Clarify BYOK / Custom Inference description copy#11780
Conversation
|
I'm starting a first review of this pull request. You can follow along in the session on Warp. I completed the review and no human review was requested for this pull request. Comment Powered by Oz |
There was a problem hiding this comment.
Overview
This PR updates the Custom inference description copy in the AI settings panel to clarify that API keys are stored locally and are not retained on Warp's servers.
Concerns
- This is a user-facing settings copy change, but the PR does not include a screenshot or screen recording demonstrating the updated text in the UI.
Verdict
Found: 0 critical, 1 important, 0 suggestions
Request changes
Comment /oz-review on this pull request to retrigger a review (up to 3 times on the same pull request).
Powered by Oz
| let text_fragments = vec![ | ||
| FormattedTextFragment::plain_text( | ||
| "Use your own API keys from model providers for Warp Agent. You can also add custom endpoints to use third-party models. Custom endpoints must support the OpenAI-compatible Chat Completions API. API keys are stored locally and are never synced to the cloud. Using auto models or models from providers you have not provided API keys for will consume Warp credits. ", | ||
| "Use your own API keys from model providers for Warp Agent. You can also add custom endpoints to use third-party models. Custom endpoints must support the OpenAI-compatible Chat Completions API. API keys are stored locally on your device and are never retained on Warp's servers. Using auto models or models from providers you have not provided API keys for will consume Warp credits. ", |
There was a problem hiding this comment.
e2cda90 to
4618506
Compare
Per Petra's review feedback on the BYOK docs (warpdotdev/docs#138) and a related Reddit thread, the current Custom inference description in the AI settings panel says API keys 'are never synced to the cloud'. That phrasing implies keys never leave the user's device — but BYOK keys do transit Warp's backend in-flight per request, they just aren't stored there. Replace with two short sentences: keep the privacy assurance ('stored only on your device, never on Warp's servers') and add a purpose statement ('They're used to make requests to your chosen model provider') so the description also conveys what the key actually does. Co-Authored-By: Oz <oz-agent@warp.dev>
4618506 to
3f4dcb8
Compare
The in-app description for the Custom inference settings panel landed on a simpler framing (warpdotdev/warp#11780): API keys are stored only on your device, never on Warp's servers. They're used to make requests to your chosen model provider. Mirror that framing as the headline statement in the docs so the docs and in-app copy use the same vocabulary. The 3-step request flow and the 'passes through but isn't stored' elaboration below remain in place to answer the deeper 'how does the key get to the provider?' question for users who want detail. Same change applied in parallel to the Custom Inference Endpoint page (both the Key features bullet and the How it works intro). Co-Authored-By: Oz <oz-agent@warp.dev>
This PR replaces the single misleading sentence with two short ones that keep the privacy assurance ("stored only on your device, never on Warp's servers") and add a purpose statement ("They're used to make requests to your chosen model provider") so the description also conveys what the key actually does.
Diff
Linked Issue
ready-to-specorready-to-implement.Pairs with warpdotdev/docs#138.
Testing
./script/runThis is a static settings-panel string change in
render_custom_inference_description(app/src/settings_view/ai_page.rs). No code paths or behavior change; only the user-facing description text in the Custom inference section.Screenshots / Videos
N/A — visual change is purely the replacement of one sentence in the Custom inference description block.
Agent Mode
CHANGELOG-BUG-FIX: Clarify BYOK / Custom inference settings copy: API keys are stored only on your device, never on Warp's servers, and used to make requests to your chosen model provider.
Conversation: https://staging.warp.dev/conversation/12c60bdb-97b9-4fe4-8c1d-7082bcb57992
Run: https://oz.staging.warp.dev/runs/019e6a71-fc28-78c8-952c-f07a9c098ab2
This PR was generated with Oz.