skills.rest
moongci38-oss
moongci38-ossCOMMUNITY
Explore all skills by this author →
GitHub

start-sonnet

Start Sonnet implementation sessions with context loading and Evaluator quality control.

Updated Jun 3, 2026
Search Tags:#implementation#context loading#session start#Sonnet#Evaluator
One-click install
npx skills add https://github.com/moongci38-oss/forge-plugins --skill start-sonnet
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: start-sonnet
Source: https://github.com/moongci38-oss/forge-plugins/tree/main/forge-core/skills/start-sonnet
Command: npx skills add https://github.com/moongci38-oss/forge-plugins --skill start-sonnet

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires claude-haiku-4-5, and includes scripts (resource) and evals (resource) components.

What problem does it solve?

Manages the start of Sonnet sessions, loading necessary implementation context efficiently.

Core Features & Use Cases

  • Session Start Trigger: Initiated by triggers such as "session start", "implementation start", "start-sonnet", and upon receiving Opus handover.
  • Evaluator: Validates output quality using a separate Evaluator subagent with the claude-haiku-4-5 model.
  • Output Quality Metrics: Assesses clarity, consistency, completeness, and safety.

Quick Start

To initiate a Sonnet session, simply execute the command /start-sonnet.

Frequently Asked Questions about start-sonnet

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I start an implementation session with Sonnet after an Opus handover?▼

Context loading for Sonnet implementation sessions is triggered by commands like 'session start' or 'start-sonnet'. It prepares the environment by loading necessary implementation context, which is especially useful when receiving an Opus handover.

How does output validation work during Sonnet session initialization?▼

Output validation during session initialization works by utilizing an Evaluator subagent powered by the claude-haiku-4-5 model. This independent subagent checks output quality metrics including clarity, consistency, completeness, and safety.

Do I need claude-haiku-4-5 to validate Sonnet implementation outputs?▼

Yes, you need claude-haiku-4-5 because it is required for the Evaluator subagent to perform independent output validation. This dependency ensures that session outputs meet quality standards during implementation.

What quality metrics are assessed when loading context for an implementation session?▼

When loading context for an implementation session, the Evaluator subagent assesses clarity, consistency, completeness, and safety. These output quality metrics ensure the session starts with validated and reliable implementation context.

Can I use session start triggers for Sonnet implementation without an Opus handover?▼

Yes, you can use triggers such as 'session start', 'implementation start', or 'start-sonnet' to initiate Sonnet implementation with context loading. Receiving an Opus handover is just one of the applicable trigger scenarios.

Related Skills

Explore More Software Engineering Skills→
moongci38-ossmoongci38-oss

start-opus

Initialize an Opus session by reading the latest handover file and loading context.

Community
Intermediate
nrwlnrwl
29.2k

run-nx-generator

Run Nx generators with priority for workspace-plugin generators in monorepos.

Official
Advanced
gopforevergopforever

ai-integrations-anthropic

Enable Anthropic chat and completions via the Replit AI Integrations proxy.

Community
Advanced
Sake-TeamSake-Team

ck

Manage persistent project context and session state using JSON.

Official
Advanced