claude-code-usage

[AUTO-INVOKE] MUST be invoked at the START of each new coding session. Covers context management, task strategies, and Foundry-specific workflows. Trigger: beginning of any new conversation or coding session in a Solidity/Foundry project.

Invalid

This skill can't be scored yet

Validation errors are blocking scoring. Review and fix them to unlock Quality, Impact and Security scores. See what needs fixing →

Evaluation results

100%

36%

Foundry Project Contribution Guide

Foundry commit and git workflow

Criteria

Without context

With context

forge fmt && forge test before commit

100%

forge build after contract edit

100%

forge test --gas-report before PR

100%

forge test -vvvv for debugging

100%

git diff before commit

70%

100%

No git add .

100%

.env commit risk

37%

100%

Stage specific files only

100%

No push to main/master

20%

100%

Commit before push rule

100%

Without context: $0.1932 · 59s · 13 turns · 19 in / 2,761 out tokens

With context: $0.1924 · 46s · 11 turns · 14 in / 2,276 out tokens

100%

67%

AI Assistant Usage Guide for Solidity Teams

Context management and meta-prompting guidance

Criteria

Without context

With context

/plan for multi-file changes

100%

/plan for large features

100%

/clear over /compact

100%

One task per window

100%

/clear after complex tasks

25%

100%

Meta-prompting: diagnose step

100%

Meta-prompting: rewrite step

100%

Meta-prompting: /clear and repaste

100%

Specific paths in prompts

100%

Set prompt boundaries

100%

No unrelated tasks in one message

100%

.claude/instructions.md config

40%

100%

Without context: $0.2522 · 1m 54s · 12 turns · 19 in / 5,336 out tokens

With context: $0.2837 · 1m 36s · 15 turns · 17 in / 4,594 out tokens

100%

ERC-20 代币合约开发

Language-adaptive response

Criteria

Without context

With context

README in Chinese

100%

Code comments in Chinese

100%

NatSpec in Chinese

100%

No English-only explanation

100%

Error/event names may stay English

100%

Summary file in Chinese

100%

Without context: $0.2063 · 1m 9s · 11 turns · 16 in / 4,245 out tokens

With context: $0.4682 · 1m 53s · 22 turns · 26 in / 6,583 out tokens

Repository: NeverSight/skills_feed
Commit: 5342bca

Evaluated: 13 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Foundry Project Contribution Guide AI Assistant Usage Guide for Solidity Teams ERC-20 代币合约开发

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.