popeye-cli 1.4.0 → 1.4.1
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 +66 -0
- package/package.json +1 -1
package/CHANGELOG.md
ADDED
|
@@ -0,0 +1,66 @@
|
|
|
1
|
+
# Changelog
|
|
2
|
+
|
|
3
|
+
All notable changes to Popeye CLI are documented in this file.
|
|
4
|
+
|
|
5
|
+
## [1.4.0] - 2026-02-08
|
|
6
|
+
|
|
7
|
+
### Added
|
|
8
|
+
- **Project Type Upgrade (`/upgrade`)** - Expand projects to new types at any time (e.g., fullstack -> all). Includes transactional upgrade with backup/rollback, automatic planning for new apps, and integration guidance for existing code.
|
|
9
|
+
- **Multi-Provider Model Switching (`/model`)** - Switch AI models per provider: `/model openai gpt-5`, `/model gemini gemini-2.5-pro`, `/model grok grok-3`. All 3 models persisted to `popeye.md` and loaded on resume.
|
|
10
|
+
- **Flexible Model Validation** - OpenAI and Gemini schemas now accept any model name (not just hardcoded enums), with known models shown as suggestions via `/model <provider> list`.
|
|
11
|
+
- **Upgrade Context Builder** - After upgrade, scans existing apps (dependencies, API routes, shared packages) and builds rich context so the planner focuses only on new apps with integration guidance.
|
|
12
|
+
- **Monorepo-Aware Codebase Analysis** - `getProjectContext()` now detects code in `apps/*/` subdirectories, not just root.
|
|
13
|
+
- `/config model` subcommand for model management.
|
|
14
|
+
- `isWorkspace()` helper replacing scattered `=== 'fullstack'` checks.
|
|
15
|
+
- `KNOWN_OPENAI_MODELS` and `KNOWN_GEMINI_MODELS` constants for display.
|
|
16
|
+
|
|
17
|
+
### Fixed
|
|
18
|
+
- **Critical**: `ProjectStateSchema` only accepted 3 of 5 language types, breaking `website` and `all` projects at the persistence layer. Now uses `OutputLanguageSchema` as single source of truth.
|
|
19
|
+
- **Critical**: `readPopeyeConfig()` used hardcoded language array, rejecting valid `website`/`all` values from `popeye.md`.
|
|
20
|
+
- `default_language` in config schema used hardcoded enum instead of `OutputLanguageSchema`.
|
|
21
|
+
- `/lang` help text now shows all 5 language shortcuts.
|
|
22
|
+
|
|
23
|
+
### Changed
|
|
24
|
+
- `OpenAIModelSchema` changed from strict `z.enum()` to flexible `z.string().min(1)`.
|
|
25
|
+
- `GeminiModelSchema` changed from strict `z.enum()` to flexible `z.string().min(1)`.
|
|
26
|
+
- `GeminiModel` type changed from union literal to `string` in both `consensus.ts` and `gemini.ts`.
|
|
27
|
+
- All `=== 'fullstack'` workspace checks replaced with `isWorkspace()` (covers both `fullstack` and `all`).
|
|
28
|
+
|
|
29
|
+
## [1.3.0] - 2026-02-05
|
|
30
|
+
|
|
31
|
+
### Added
|
|
32
|
+
- Website app support for ALL project type.
|
|
33
|
+
- Module exports for generator functions.
|
|
34
|
+
|
|
35
|
+
## [1.2.1] - 2026-02-05
|
|
36
|
+
|
|
37
|
+
### Fixed
|
|
38
|
+
- Rate limit handling with graceful pause instead of failure.
|
|
39
|
+
- Workflow saves progress on rate limit so `/resume` continues from where it stopped.
|
|
40
|
+
|
|
41
|
+
## [1.2.0] - 2026-02-04
|
|
42
|
+
|
|
43
|
+
### Added
|
|
44
|
+
- Grok AI provider support (reviewer/arbitrator).
|
|
45
|
+
- Fullstack project type with frontend + backend monorepo.
|
|
46
|
+
- `popeye.md` project configuration file.
|
|
47
|
+
- Multi-provider consensus (OpenAI + Gemini + Grok).
|
|
48
|
+
|
|
49
|
+
## [1.1.0] - 2026-02-03
|
|
50
|
+
|
|
51
|
+
### Added
|
|
52
|
+
- npm version badge and quick install instructions.
|
|
53
|
+
- CONTRIBUTING.md guidelines.
|
|
54
|
+
- Open Source Manifesto.
|
|
55
|
+
|
|
56
|
+
## [1.0.0] - 2026-02-01
|
|
57
|
+
|
|
58
|
+
### Added
|
|
59
|
+
- Initial release: AI-powered autonomous code generation.
|
|
60
|
+
- Interactive Claude Code-style CLI interface.
|
|
61
|
+
- Plan mode with idea expansion and consensus review.
|
|
62
|
+
- Execution mode with per-task planning, consensus, and implementation.
|
|
63
|
+
- Python and TypeScript project generators.
|
|
64
|
+
- OpenAI and Gemini adapter support.
|
|
65
|
+
- Test runner with automatic retry.
|
|
66
|
+
- Project state persistence and resume.
|