| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| README.md | 2026-01-05 | 1.4 kB | |
| v1.4.2 - Model Selection _ npx Portability source code.tar.gz | 2026-01-05 | 825.7 kB | |
| v1.4.2 - Model Selection _ npx Portability source code.zip | 2026-01-05 | 946.4 kB | |
| Totals: 3 Items | 1.8 MB | 0 | |
Model Selection for Orchestration Skill
Every Task() call in the orchestration skill now includes an explicit model= parameter with clear guidance on when to use each tier.
Model Tier Framework
| Model | Best For |
|---|---|
| haiku | Exploration, searching, fetching context, data discovery - spawn many |
| sonnet | Implementation, test generation, documentation, structured work |
| opus | Code review, security analysis, architecture, conflict resolution |
The Mental Model
- Haiku = "errand runner" — spawn wildly, gather info, no judgment
- Sonnet = "junior-mid engineer" — well-structured tasks, clear direction needed
- Opus = "critical thinker" — trust its judgment, ambiguous problems, architecture
All Commands Now Portable with npx
All command examples across the codebase now use npx cc-mirror instead of cc-mirror for maximum portability - no global install required!
Updated guidance in: - Task Manager skill - README.md - Team Mode docs - CLI help text - AGENTS.md - Architecture docs - Issue templates
Changes
- Added Model Selection section to
SKILL.mdwith tier descriptions - Updated 12 skill reference files with explicit
model=parameters - Updated all CLI command examples to use
npx cc-mirror - 22 files updated for consistency
Full Changelog: https://github.com/numman-ali/cc-mirror/compare/v1.4.0...v1.4.2