thegraph

I took a look at your thegraph skill and wanted to share some thoughts.

Links:

• GitHub
• Your skill in the SkillzWave marketplace

The TL;DR

You're at 81/100, solidly in B territory. This is based on Anthropic's skill best practices rubric. Your strongest area is Ease of Use (22/25) – the trigger patterns and metadata are excellent. The biggest drag is Progressive Disclosure Architecture (24/30) – you've got ~230 lines without a table of contents, and some MCP code blocks are doing heavy lifting when they could be leaner.

What's Working Well

• Excellent trigger coverage – You're catching schema.graphql files, @entity patterns, and AssemblyScript handlers. That's the kind of specificity that makes a skill actually discoverable.
• Strong anti-patterns section – Showing what NOT to do (infinite loops, incorrect IDs) is more useful than just examples. Developers learn faster from "don't do this" paired with "do this instead."
• Complete walkthrough examples – You've got a real schema + handler combo that shows the whole flow. Not abstract, not pseudo-code – actual TheGraph patterns.
• Practical constraints – The banned AssemblyScript operations and validation rules aren't pedantic; they address real foot-guns in subgraph development.

The Big One: Navigation for Dense Content

Your file is 230+ lines covering a bounded but complex domain. Without a table of contents, a developer using an AI agent (which is your actual user) has to parse through linearly. The XML section tags help, but they're not indexed.

Fix: Add a TOC after your objective tag:

## Sections
- [Quick Start](#quick_start)
- [AssemblyScript Rules](#assemblyscript_rules)
- [MCP Query Patterns](#mcp_first)
- [Schema & Handler Example](#complete_example)
- [Success Criteri...