command-code 1.72.4 → 1.73.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/CHANGELOG.md CHANGED
@@ -1,5 +1,18 @@
1
1
  # command-code
2
2
 
3
+ ## 1.73.1
4
+
5
+ ### Patch Changes
6
+
7
+ - feat: retire Pixel Canary when its stealth preview ends on September 30, 2026 at 11 PM Pacific
8
+ - fix: `taste push --all --merge` now merges into the remote project instead of silently overwriting it
9
+
10
+ ## 1.73.0
11
+
12
+ ### Minor Changes
13
+
14
+ - feat: Select separate models for planning and implementation in `/config`
15
+
3
16
  ## 1.72.4
4
17
 
5
18
  - chore: queue support in new experimental features
@@ -19,6 +19,8 @@ The executable registry is authoritative. Run `cmd config list --scope effective
19
19
  | `taste-learning` | user | boolean | `true`, `false` | Learn your coding style from your sessions, user-wide (same setting as /taste). |
20
20
  | `show-taste-learning-in-feed` | user | boolean | `true`, `false` | Show Taste learning notices in the conversation feed. Turning this off keeps learning enabled and only hides the notices. |
21
21
  | `taste-learning-project` | project | enum | `inherit`, `enabled`, `disabled` | Learn your coding style from this project only (same setting as /taste). |
22
+ | `feature-model:planning` | user | enum | known model id or `default` | Model that runs while you are in plan mode (shift+tab, /plan, or enter_plan_mode) — exploring the codebase and writing the plan. Value: a model id, or "default" to reset so the phase follows your current model. |
23
+ | `feature-model:implementation` | user | enum | known model id or `default` | Model that takes over when you approve a plan — writing the code. Only applies after a planning round. Value: a model id, or "default" to reset so the phase follows your current model. |
22
24
  | `feature-model:titleGeneration` | user | enum | known model id or `default` | Names your sessions after the first exchange. Value: a model id, or "default" to reset to the curated default. |
23
25
  | `feature-model:compaction` | user | enum | known model id or `default` | Summarizes long conversations when context fills up. Value: a model id, or "default" to reset to the curated default. |
24
26
  | `feature-model:toolDescription` | user | enum | known model id or `default` | Writes the one-line command summaries on permission prompts. Value: a model id, or "default" to reset to the curated default. |
@@ -108,7 +108,7 @@ Fires after the turn commits. **Any** mod returning `true` ends the run with `st
108
108
 
109
109
  ### `prepareNextTurn({state, turnNumber}, ctx?) → {model?, effort?} | undefined`
110
110
 
111
- Fires at the bottom of a continuing round. Results are merged across mods - later mods override earlier ones per field. A returned `model`/`effort` applies from the **next** model call onward (`config` is never mutated).
111
+ Fires at the bottom of a continuing round. Results are merged across mods - later mods override earlier ones per field. A returned `model`/`effort` applies from the **next** model call onward (`config` is never mutated). `effort` is tri-state: a `ThinkingEffort` sets it, `null` **resets** it to the incoming model's own default, and omitting the field leaves the current effort unchanged. Send `null` (not an omitted field) when swapping to a model that should run at its own default - otherwise the run keeps the outgoing model's effort.
112
112
 
113
113
  ### `appendSystemPrompt({state}, ctx?) → string | undefined`
114
114