Review - #144
Draft
quangdang46 wants to merge 17 commits into
Draft
Conversation
…g-agent CLI repos - Introduced a new Bash script `clone_repo.sh` to clone coding-agent CLI repositories into a specified directory. - Enhanced the existing `clone_repo.bat` script with improved comments and additional repository URLs for cloning. - Both scripts now provide clearer instructions and feedback during the cloning process.
- Deleted the JetBrains/junie repository cloning command from both `clone_repo.bat` and `clone_repo.sh` scripts to streamline the cloning process and remove unnecessary dependencies.
install_release.sh and rebuild.sh wrote `ln -sfn "next-code"` onto the
launcher path itself, producing a self-referential symlink ("too many
levels of symbolic links") that broke the launcher, setup-hotkey, and
server reload after every install. install.sh still referenced the
undefined $compat_launcher_path after the jcode purge. Drop all compat
stubs since the launcher is already named next-code.
The rebrand's one-release compat was `jcode -> next-code`; the purge commit rewrote it onto the launcher's own path, making it a self-referential link (and leaving $compat_launcher_path undefined in install.sh / a self-copy in install.ps1). Restore the compat entry under the no-dash name `nextcode` (which Cargo.toml also ships as a bin target), so `nextcode` works alongside `next-code`.
…ect reasoning - Add GPT-5.6 model family (sol/pro/standard/terra/luna) incl. ChatGPT web route gpt-5.6-pro[web] and API-key-only Pro model helper. - Register Meta Muse (META_MUSE_API_KEY) OpenAI-compatible profile + login provider; bump profile/login catalog sizes. - Keep direct DeepSeek routes emitting reasoning_content (issue #815).
Wire the previously-dead is_openai_api_only_pro_model helper into route building, availability, and the OpenAI runtime so the 5 GPT Pro models (gpt-5.6/5.5/5.4/5.2/5-pro) are never selectable or routable via the ChatGPT/Codex OAuth route. The platform API-key route (openai-api-key) keeps them. - provider-core: add is_openai_api_only_pro_model_for_route predicate + re-export; spec-parser test documents OAuth specs still parse (route stays visible-but-unavailable). - base: OAuth route builders emit Pro models as visible-but-unavailable with a platform-key detail; model_availability_for_account returns Unavailable for Pro without an API key; get_best_available_openai_model skips Pro on OAuth-only creds. - openai-runtime: set_model bails on Pro + ChatGPT/Codex-shaped creds; model_id auto-fallback never lands an OAuth session onto a Pro model. - tests: predicate, route-builders, availability, and set_model gates.
feat(providers): GPT-5.6 family, Meta Muse provider, DeepSeek direct reasoning
feat(providers): gate GPT-5.x Pro models behind OpenAI platform API key
Mirror the seven_day fallback dedup onto the five_hour fallback: when a weekly-named window is classified as the seven_day window, do not also grab it as the generic five_hour window. Previously a single 'Codex weekly' limit could be counted in both windows.
Port jcode's todo status fixes (jcode commits 1ab482c77 + d1f24ad7f): - Add canonical_todo_status helper mapping natural/case-varied synonyms (done/finished/complete -> completed, canceled -> cancelled). - normalize_todo_input now canonicalizes model-written statuses. - The todo tool rejects unknown status values with a clear vocabulary message, and its schema advertises the enum [pending, in_progress, completed, cancelled]. - Auto-poke / group-complete / remaining-count checks use the canonical helpers so persisted 'done'/'finished' sessions are treated as complete.
qdang46
marked this pull request as draft
August 12, 2026 03:32
Bring jcode's Grok Build ACP provider into next-code as a CLI/TUI provider:
next-code spawns the Grok CLI ('grok agent stdio') and drives it as the
coding agent over ACP JSON-RPC, streaming ThinkingDelta/TextDelta back.
No xAI HTTP / API-key path — auth is the Grok CLI's own cached login.
- New crate next-code-provider-grok-runtime: GrokBuildProvider (ACP client
over stdio), fake-ACP test binary, and 3 integration tests (handshake +
model prefetch + auth isolation, resume without history replay, drop
cancels the prompt).
- Auth: grok_build module (cli_path/cli_available), AuthStatus.grok_build
field, state/assessment arms (LocalCliSession, CommandProbe).
- Metadata: GROK_BUILD_LOGIN_PROVIDER (Cli auth kind), LoginProviderTarget/
AuthStateKey variants.
- Provider layer: ActiveProvider::GrokBuild, RuntimeProviderId::GrokBuild,
RuntimeKey/ModelRouteApiMethod::GrokBuild, dispatch + set_model +
hot-init arms, ProviderAvailability.grok_build.
- CLI: ProviderChoice::GrokBuild, login_grok_build_flow (runs 'grok login'),
bootstrap via external GROK_RUNTIME registry, report_info/auth_test
coverage, startup registration + test.
- Legacy TUI: GrokBuild login arm (plain 'run grok login' notice only);
no deeper legacy-TUI integration (grok-tui is the main UI).
Env override: NEXT_CODE_GROK_CLI_PATH.
fix(usage): avoid duplicate Codex quota windows (fixes #869)
…ject fix(todo): reject unknown statuses and normalize completion synonyms
feat(providers): add Grok Build ACP provider
…lished beads - AGENTS.md: add bv-agent-instructions-v3 (beads workflow integration block). - .beads/issues.jsonl: 24 polished beads (4 P0/15 P1/3 P2/1 P3) covering audit findings R1-R20 and parity features NX-*, with dependencies and full self-contained specs.
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
2 participants
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.
No description provided.