Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/thesethrose/copilot-skills/mcp-buildergit clone --depth 1 https://github.com/TheSethRose/Copilot-SkillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/thesethrose/copilot-skills/mcp-builder)<a href="https://agentmods.dev/instructions/thesethrose/copilot-skills/mcp-builder"><img src="https://agentmods.dev/badge/instructions/thesethrose/copilot-skills/mcp-builder.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.01705 | $0.01705 |
| Opus 5 | $0.00852 | $0.00852 |
| Sonnet 5 | $0.00341 | $0.00341 |
| Haiku 4.5 | $0.00170 | $0.00170 |
Grade A, and why
Copilot-Skills mcp-builder.instructions.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 278 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MCP Builder - Automatic Context Instructions
Related Prompt: /mcp-builder
Related Skill: .github/copilot-skills/mcp-builder/README.md
Triggers: mcp, model context protocol, mcp server, agent tools, llm integration, api integration, fastmcp, tool design
Context: MCP Server Development
When working with MCP-related files or when user queries contain MCP server keywords, this context is automatically activated.
Default Behaviors
When user mentions "mcp server" or "build mcp"
- Suggest
/mcp-builderskill prompt - Load agent-centric design principles
- Reference framework-specific documentation
- Use evaluation-driven development approach
When user mentions "tool design" or "agent tools"
- Focus on workflow-oriented tools (not just API wrappers)
- Optimize for context efficiency
- Design actionable error messages
- Follow natural task subdivisions
When user mentions "python mcp" or "fastmcp"
- Load Python-specific patterns
- Reference FastMCP or Python SDK documentation
- Use Pydantic for input validation
- Follow Python project structure conventions
When user mentions "typescript mcp" or "node mcp"
- Load TypeScript-specific patterns
- Reference MCP TypeScript SDK documentation
- Use Zod for input validation
- Follow Node/TypeScript project conventions
Quality Guidelines
✅ Do
- Build workflow tools, not just API endpoint wrappers
- Consolidate related operations into single tools
- Return high-signal information (concise by default)
- Use human-readable identifiers (names over IDs)
- Design actionable, educational error messages
- Validate all inputs with schemas (Pydantic/Zod)
- Handle rate limiting and pagination gracefully
- Test with realistic agent scenarios
- Provide configurable detail levels (concise/detailed)
- Keep responses under 25,000 tokens
- Group related tools with consistent prefixes
❌ Don't
- Simply wrap API endpoints without workflow consideration
- Return exhaustive data dumps (respect context limits)
- Use cryptic error messages
- Skip input validation
- Ignore rate limits or pagination
- Hard-code configuration values
- Mix workflow concerns in single tool
- Create tools with unclear purposes
- Return massive unformatted responses
- Use technical IDs when names are available
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 278 lines · 1,705 tokens per session scan A 26435b798042
Copilot-Skills mcp-builder.instructions.md is an instructions file published in the GitHub repository TheSethRose/Copilot-Skills (3 stars, last pushed 9mo ago), licensed MIT. It adds 1,705 tokens to every session, about $0.0085 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
turbo CLAUDE.md
Claude Code instructions for tobihagemann/turbo, covering turbo, project structure and key files.
platform-skills copilot-instructions.md
Copilot instructions for nitinjain999/platform-skills, covering platform engineering instructions for github copilot, version: 1.40.0, source: https://github.com/nitinjain999/platform-skills, response contract and core principles.
claude-code-toolkit CLAUDE.md
Claude Code instructions for robertguss/claude-code-toolkit, covering claude.md, project overview, build commands, package a single skill for claude.ai upload and package all skills.
agent-teacher CLAUDE.md
Claude Code instructions for JackyYang258/agent-teacher, covering agentteacher, before you start, common commands and project-specific hard rules.
github-copilot-superpowers subagents.instructions.md
Instructions for kasuken/github-copilot-superpowers, covering subagent instructions (ssot), terminology, agent role: orchestrator only, ⚠️ absolute rules (non-negotiable) and mcp delegation rules.
multai CLAUDE.md
Claude Code instructions for alo-exp/multai, covering multai — claude code instructions, 0. session startup (automatic), project overview, 1. automated enforcement and 2. active workflow.