Templates

Skill Benknightdark/neo-skills/.agents/skills/generate-skill/templates

Neo Skills 是專為現代 AI Agent 設計的全方位能力擴充套件。本專案透過標準化的通訊架構，為 AI 代理安裝可插拔的「技能模組 (Skills)」，使其不僅僅是一個聊天機器人，而是能轉化為具備「感知-推理-行動」能力的多領域專家。

Install

npx -y skills add Benknightdark/neo-skills --skill templates

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

7 stars7 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use this skill when the user wants to [intended task / user intent]. It triggers when they mention [keywords / concepts / adjacent tasks], even if they do not explicitly say "[core domain]".

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

2.4 KB, as published. Nobody here has run it

Skill Name

Briefly summarize the target task, scope, and expected outcome of this skill.

Gotchas

Gotcha 1: [Describe a highly specific, non-obvious, environment-dependent pitfall that models are likely to hit, rather than generic advice].
Gotcha 2: [e.g., database soft-deletes, multi-system field name mapping, or silent health checks].

Workflow Checklist

Progress:

Step 1: Parse and Inventory (Load references/schema.md or execute scripts/parse.py).
Step 2: Formulate Plan (Create a plan JSON and validate using scripts/validate.py).
Step 3: Execute Task (Apply changes with idempotent actions).
Step 4: Verify Results (Run tests and confirm output formatting).

Detailed Guidelines

Step 1 — Parse & Inventory

Instruct the agent how to analyze the environment:

Load references/guidelines.md for specific style rules.
Verify input arguments. If ambiguous, reject immediately with clear options.

Step 2 — Validation Loop

Instruct the agent to execute a deterministic loop to prevent broken output:

Perform the edit.

Validate using a validation script:

python3 scripts/validate.py "$TARGET_FILE"

If validation fails:
- Carefully inspect error codes and diagnostics in stderr.
- Revise target code/text.
- Run validation again.
Do NOT proceed to the next step until validation completely passes.

Step 3 — Plan-Validate-Execute

For batch, risky, or destructive actions:

Extract data structures into a planning file plan.json.
Compare plan fields against the schema.

Validate:

python3 scripts/compare_schema.py plan.json references/schema.json

If valid, proceed with --confirm or --force.

Output Templates

When presenting results, format exactly like this template:

# [Task Title]

## Executive Summary
[Brief description of changes made]

## Applied Checklist
- [x] Change 1
- [x] Change 2

## Recommendations / Next Steps
1. [First actionable step]
2. [Second actionable step]