feat(voice): v0.0.13 — ChatGPT Voice catalog discovery (list_voices) - #31
Open
robotlearning123 wants to merge 12 commits into
Open
feat(voice): v0.0.13 — ChatGPT Voice catalog discovery (list_voices)#31robotlearning123 wants to merge 12 commits into
robotlearning123 wants to merge 12 commits into
Conversation
Reject a voices list above _MAX_VOICES (128) as contract drift instead of normalizing an unbounded private response, and add tests at and above the bound. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01VAbFRELj9M3MYQt7SkUaLG
Document list_voices as the 26th tool across README, docs, skill, and reference; state that catalog discovery is supported while GPT-Live realtime audio, transport, synthesis, and transcript guarantees are not. Add docs/roadmap.md (version lanes, GPT-Live boundary, language policy, release gates) and an install regression asserting the bundled skill allowlists all 26 tools. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01VAbFRELj9M3MYQt7SkUaLG
Bump pyproject, package, plugin, and server metadata to 0.0.13 and add the dated CHANGELOG section for the read-only Voice catalog. Release is held (no tag or publish) until 0.0.12 lands on its own lane. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01VAbFRELj9M3MYQt7SkUaLG
Forward an optional voice_mode query param (standard/advanced/live/wingman and future modes) to /backend-api/settings/voices; charset-validate it and reject malformed values before any request. Cover mode passthrough, default omission, and rejection offline, and parametrize the opt-in live contract test over the real modes. list_voices(voice_mode="live") targets the GPT-Live catalog. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01VAbFRELj9M3MYQt7SkUaLG
📝 WalkthroughWalkthroughAdds a read-only ChangesVoice catalog release
Estimated code review effort: 3 (Moderate) | ~30 minutes Possibly related PRs
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 3
🧹 Nitpick comments (1)
gpt2agent/tools/voice.py (1)
29-31: 📐 Maintainability & Code Quality | 🔵 Trivial | 💤 Low valueConsider typing
_fail_contractasNoReturn.The function always raises but is annotated
-> None. UsingNoReturnwould let type checkers narrow types correctly in_bounded_textand_normalize_catalog(e.g., confirmingvalueisstrafter the guard).♻️ Optional refactor
-from typing import Any +from typing import Any, NoReturn -def _fail_contract() -> None: +def _fail_contract() -> NoReturn: """Raise a payload-free error for a private response-shape change.""" raise RuntimeError(_CONTRACT_ERROR)🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@gpt2agent/tools/voice.py` around lines 29 - 31, Update the return annotation of _fail_contract from None to NoReturn and ensure NoReturn is imported from the typing module as needed. Leave its existing RuntimeError behavior unchanged so type checkers can narrow guarded values in _bounded_text and _normalize_catalog.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/roadmap.md`:
- Line 22: In the roadmap wording near the Voice product note, replace the
phrase “post-merge main checks” with “post-merge checks on `main`,” preserving
the surrounding text.
In `@docs/superpowers/plans/2026-07-10-v0.0.13-voice-release.md`:
- Around line 65-75: Update the voice-listing contract and implementation around
_normalize_catalog and list_voices to accept an optional voice_mode parameter,
omit it when unset, and URL-encode it in the request query when provided.
Validate supported lowercase mode values and explicitly reject live; add
coverage for omitted mode, supported modes, query encoding, invalid values, and
live.
In `@docs/superpowers/specs/2026-07-10-account-native-feature-coverage-design.md`:
- Around line 258-263: Update the Voice API records to document the implemented
voice_mode contract: identify the accepted modes and explicitly state that live
is rejected. Apply this clarification in the list_voices() section of
docs/superpowers/specs/2026-07-10-account-native-feature-coverage-design.md
(lines 258-263) and the corresponding review conclusions in
docs/superpowers/reviews/2026-07-10-account-native-feature-coverage-cross-model-review.md
(lines 67-76), keeping both descriptions aligned.
---
Nitpick comments:
In `@gpt2agent/tools/voice.py`:
- Around line 29-31: Update the return annotation of _fail_contract from None to
NoReturn and ensure NoReturn is imported from the typing module as needed. Leave
its existing RuntimeError behavior unchanged so type checkers can narrow guarded
values in _bounded_text and _normalize_catalog.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 6a900e60-bcd0-4dce-ab4e-9de6416bbc72
📒 Files selected for processing (26)
.claude-plugin/marketplace.json.claude-plugin/plugin.jsonCHANGELOG.mdCLAUDE.mdCONTRIBUTING.mdREADME.mddocs/README.mddocs/faq.mddocs/how-it-works.mddocs/roadmap.mddocs/superpowers/plans/2026-07-10-v0.0.13-voice-release.mddocs/superpowers/reviews/2026-07-10-account-native-feature-coverage-cross-model-review.mddocs/superpowers/specs/2026-07-10-account-native-feature-coverage-design.mdgpt2agent/__init__.pygpt2agent/install.pygpt2agent/skills/gpt2agent/SKILL.mdgpt2agent/skills/gpt2agent/tools-reference.mdgpt2agent/tools/__init__.pygpt2agent/tools/voice.pypyproject.tomlserver.jsontests/test_audit_2026_07_09_tools.pytests/test_backend_tools.pytests/test_install.pytests/test_none_guards.pytests/test_tools.py
|
|
||
| ## The GPT-Live boundary | ||
|
|
||
| Voice is an official ChatGPT product, but the routes this project touches are |
There was a problem hiding this comment.
📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win
Clarify the main branch gate wording.
Replace “post-merge main checks” with “post-merge checks on main” to avoid ambiguity.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@docs/roadmap.md` at line 22, In the roadmap wording near the Voice product
note, replace the phrase “post-merge main checks” with “post-merge checks on
`main`,” preserving the surrounding text.
Source: Linters/SAST tools
Comment on lines
+65
to
+75
| - [ ] Implement a pure `_normalize_catalog(data)` helper with these rules: | ||
| - require a dict envelope and list-valued `voices`; | ||
| - accept an empty list as an honest empty catalog; | ||
| - require every item to be a dict with printable, non-empty string `voice`, `name`, and `description` fields, bounded to 128, 256, and 2,000 characters respectively; | ||
| - require `preview_url`, when present, to be a string or null; | ||
| - redact display text before returning it; | ||
| - represent missing, mistyped, or catalog-unknown selection as `selected: None`; only a selected ID present in the normalized catalog produces per-item `true/false` values; | ||
| - return only `id`, `name`, `description`, `selected`, and `has_preview`; | ||
| - raise one safe, response-free `RuntimeError` on contract drift. | ||
| - [ ] Implement async `list_voices()` using `async_get(client, "/backend-api/settings/voices", target_path="/backend-api/settings/voices")`. | ||
| - [ ] Decorate it with official MCP annotations: read-only, non-destructive, idempotent, and open-world. |
There was a problem hiding this comment.
🎯 Functional Correctness | 🟠 Major | ⚡ Quick win
Add voice_mode and explicit live rejection to the implementation contract.
The plan omits the optional voice_mode parameter, its query encoding, and validation tests. The supplied gpt2agent/tools/voice.py implementation accepts generic lowercase tokens, so live must be explicitly rejected to match the PR objective. Add tests for omitted mode, supported modes, URL encoding, invalid values, and live.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@docs/superpowers/plans/2026-07-10-v0.0.13-voice-release.md` around lines 65 -
75, Update the voice-listing contract and implementation around
_normalize_catalog and list_voices to accept an optional voice_mode parameter,
omit it when unset, and URL-encode it in the request query when provided.
Validate supported lowercase mode values and explicitly reject live; add
coverage for omitted mode, supported modes, query encoding, invalid values, and
live.
Comment on lines
+258
to
+263
| #### `list_voices()` | ||
|
|
||
| - Route: `GET /backend-api/settings/voices` | ||
| - Return live voice IDs and display metadata supplied by the account response. | ||
| - Do not hard-code documentation names because voice catalogs can be rollout-specific. | ||
| - This is catalog access only; it does not claim audio streaming or speech synthesis. |
There was a problem hiding this comment.
📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win
Keep the design and review records aligned with the shipped Voice API.
Both records describe catalog-only Voice support but omit the implemented voice_mode input contract.
docs/superpowers/specs/2026-07-10-account-native-feature-coverage-design.md#L258-L263: document accepted modes andliverejection.docs/superpowers/reviews/2026-07-10-account-native-feature-coverage-cross-model-review.md#L67-L76: record the same validation boundary in the review conclusions.
📍 Affects 2 files
docs/superpowers/specs/2026-07-10-account-native-feature-coverage-design.md#L258-L263(this comment)docs/superpowers/reviews/2026-07-10-account-native-feature-coverage-cross-model-review.md#L67-L76
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@docs/superpowers/specs/2026-07-10-account-native-feature-coverage-design.md`
around lines 258 - 263, Update the Voice API records to document the implemented
voice_mode contract: identify the accepted modes and explicitly state that live
is rejected. Apply this clarification in the list_voices() section of
docs/superpowers/specs/2026-07-10-account-native-feature-coverage-design.md
(lines 258-263) and the corresponding review conclusions in
docs/superpowers/reviews/2026-07-10-account-native-feature-coverage-cross-model-review.md
(lines 67-76), keeping both descriptions aligned.
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.
v0.0.13 — ChatGPT Voice catalog discovery
Adds
list_voices, a read-only MCP tool exposing the signed-in account's Voice catalog from the privateGET /backend-api/settings/voicesroute. Brings the server to 26 MCP tools.What's in it
list_voices— bounded, stable per-voice shape (id,name,description,selected,has_preview); backend voice IDs preserved verbatim, display text redacted.list_voices(voice_mode=...)— optional mode-specific catalog (standard/advanced/wingmanaccepted by the live contract on 2026-07-11); charset-validated before any request.voice_mode=liveis rejected by the catalog route (HTTP 422) — GPT-Live is a separate session contract, out of scope here.docs/roadmap.md— version lanes, GPT-Live boundary, language policy (Python core + optional TS sidecar), release gates.Scope guard
Catalog discovery only. Does not start a Voice session, fetch preview media, capture a microphone, synthesize speech, stream GPT-Live audio, or extract transcripts. Malformed responses fail closed (
voice catalog contract changed) rather than pretending the catalog is empty.Verification
pytest: 360 passed, 15 skipped (skips areSKIP_LIVE-gated live/network tests; no credentials configured).ruff check gpt2agent tests scripts: clean (mirrors CI).scripts/verify_release.py:release metadata verified: 0.0.13.Merge ordering
Per the roadmap, this release is held until v0.0.12 (#30) lands. This branch is based on current
mainand is independent of #30's 47 commits — after #30 merges,mainshould be merged in before release so 0.0.12's changes are carried forward.🤖 Generated with Claude Code
Summary by CodeRabbit
New Features
Documentation
Chores