Conversation
Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>
Add tab-specific design convention rules for each TUI tab (models, agents, benchmarks, status) and a CLI style guide covering colors, picker infrastructure, and output formatting. Clean up the shared TUI style guide: remove resolved inconsistencies section, fix help popup docs, add search scope descriptions, note ScrollablePanel styled title support. Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>
Add CLI interface for provider health monitoring: - `status list` — interactive picker (TTY) or table (non-TTY) - `status show <provider>` — detailed provider status view - `status status` — quick summary table of tracked providers - `status sources` — interactive picker to manage tracked providers - All commands support --json output - Respects tracked provider config (only shows tracked providers) - Reuses existing StatusFetcher pipeline (parallel fetching) Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: af08ee7fb4
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".
| #[command(name = "status")] | ||
| #[command(about = "Check AI provider service health")] |
There was a problem hiding this comment.
status entrypoint
This parser is named and documented as a standalone status command, but main still only special-cases agents and benchmarks (src/main.rs:178-193) and models link still installs only those two symlinks (src/cli/link.rs:5-32). If a user follows these examples or manually symlinks the binary as status, status list is parsed as the top-level models list command because the first real arg is just list, so the new status CLI is unreachable outside models status ....
Useful? React with 👍 / 👎.
| }) | ||
| .collect(); | ||
|
|
||
| if let Some(updated) = run_source_picker(&items)? { |
There was a problem hiding this comment.
status sources
In a non-TTY context such as models status sources | cat or CI output capture, this branch still drops into run_source_picker(), which always constructs PickerTerminal and enables raw mode (src/cli/picker.rs:29-37). Unlike status list and agents list-sources, there is no styles::is_tty() guard here, so the new command fails instead of degrading cleanly when stdout is redirected.
Useful? React with 👍 / 👎.
| components: Vec<ComponentJson<'a>>, | ||
| incidents: Vec<IncidentJson<'a>>, | ||
| scheduled_maintenances: Vec<MaintenanceJson<'a>>, |
There was a problem hiding this comment.
status show --json
For providers whose upstream only exposes summary data, or when a detail fetch fails, ProviderStatus uses the parallel components_state, incidents_state, and scheduled_maintenances_state fields to distinguish NoneReported from Unsupported/FetchFailed (src/status/types.rs:554-559). This JSON shape drops those fields, so models status show --json can only emit empty arrays, which makes downstream tooling silently interpret “detail unavailable” as “no incidents/components/maintenance”.
Useful? React with 👍 / 👎.
Add user-configurable alias names for symlink commands via [aliases] section in config.toml. Default aliases: agents, benchmarks, mstatus. Users can override any alias name in their config. Also adds status as a third symlink target (default name mstatus to avoid collision with system commands). - Add AliasesConfig struct with serde defaults for backward compat - Add AliasKind enum for routing, match_alias() and alias_names() helpers - Refactor link.rs to read alias names from config instead of hardcoded const - Update main.rs argv[0] detection to use config-driven matching - Add cli::status::run() for standalone status alias invocation - Fix link canonicalization bug (PR #28 review) - Fix link removal safety (verify symlink target before removing) - Add non-TTY fallback for status sources command (PR #29 review) - Add detail state metadata to status show --json output (PR #29 review) Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>
feat: add status CLI commands
Add user-configurable alias names for symlink commands via [aliases] section in config.toml. Default aliases: agents, benchmarks, mstatus. Users can override any alias name in their config. Also adds status as a third symlink target (default name mstatus to avoid collision with system commands). - Add AliasesConfig struct with serde defaults for backward compat - Add AliasKind enum for routing, match_alias() and alias_names() helpers - Refactor link.rs to read alias names from config instead of hardcoded const - Update main.rs argv[0] detection to use config-driven matching - Add cli::status::run() for standalone status alias invocation - Fix link canonicalization bug (PR #28 review) - Fix link removal safety (verify symlink target before removing) - Add non-TTY fallback for status sources command (PR #29 review) - Add detail state metadata to status show --json output (PR #29 review) Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>
1 participant
Summary
models status list— interactive picker (TTY) or comfy-table (non-TTY) showing tracked provider healthmodels status show <provider>— detailed provider status with components, incidents, and maintenancemodels status status— quick summary table of tracked providers (always table, even on TTY)models status sources— interactive picker to manage which providers to track (persists to config)--jsonoutputStatusFetcherpipeline with parallel fetching (10 concurrent)Test plan
models status list— verify interactive picker shows only tracked providersmodels status list --json— verify JSON outputmodels status show openai— verify detailed output with components/incidentsmodels status show openai --json— verify JSON detail outputmodels status status— verify summary tablemodels status sources— verify picker toggles and persists to configmodels status sources --json— verify source listingmodels status list | cat— verify table outputmise run fmt && mise run clippy && mise run test— all passing (264 tests)🤖 Generated with Claude Code