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/mohibshaikh/mcp-uplift/agents-mdgit clone --depth 1 https://github.com/MohibShaikh/mcp-upliftWrote 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/mohibshaikh/mcp-uplift/agents-md)<a href="https://agentmods.dev/instructions/mohibshaikh/mcp-uplift/agents-md"><img src="https://agentmods.dev/badge/instructions/mohibshaikh/mcp-uplift/agents-md.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 | $0.01748 | $0.01748 |
| Opus 5 | $0.00874 | $0.00874 |
| Sonnet 5 | $0.00350 | $0.00350 |
| Haiku 4.5 | $0.00175 | $0.00175 |
Grade A, and why
mcp-uplift AGENTS.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 5d 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 — 147 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Instructions for coding agents working in this repository. Read this before
touching anything. CLAUDE.md holds the same project conventions in more
detail; this file is the operating contract.
What this project is
mcp-uplift wraps one trusted legacy MCP stdio server and presents it behind
the stateless 2026-07-28 protocol. It is a protocol translator. It is not a
framework, not a sandbox, and not a general proxy.
It ships with zero runtime dependencies and 6 source files. That is a deliberate design property, not an accident waiting to be fixed.
Hard rules
These are not preferences. Violating any of them makes the change unacceptable regardless of how good the code is.
-
Never invent protocol behavior. Do not write a method name, field name, or error code from memory. Fetch the real thing first:
- Prose:
docs/specification/2026-07-28/in themodelcontextprotocol/modelcontextprotocolrepo (gh apiworks). - Schema and canonical payloads:
schema/2026-07-28/, especiallyexamples/, which contains real request and response JSON.
If you cannot verify a detail, stop and say so. A plausible guess about a wire format is worse than an unfinished task.
- Prose:
-
Add no dependencies. Not to
dependencies, not todevDependencies. The test suite isnode --test. If you want a library, you do not need it. -
Create no new files unless the task explicitly asks for one. Put the change where the related code already lives.
-
Do not refactor code you were not asked to change. Renaming, reformatting, reorganizing, "modernizing", extracting helpers, or tidying adjacent code are all out of scope by default. A diff that touches files unrelated to the task will be rejected.
-
Simple implementations for complex concepts. A plain
Mapand a short function beat a class hierarchy. If a concept is hard, spend the words in a comment explaining why, not in machinery. -
Comments explain why, not what. Match the existing style in
src/: every comment states the protocol constraint or the failure that forced the code to be the way it is. Never narrate what the next line already says.
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.
- 5d ago First seen · 147 lines · 1,748 tokens per session scan A 00257ba039a6
mcp-uplift AGENTS.md is an instructions file published in the GitHub repository MohibShaikh/mcp-uplift (2 stars, last pushed 20d ago), licensed MIT. It adds 1,748 tokens to every session, about $0.0087 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
helio AGENTS.md
AGENTS.md instructions for gethelio/helio, covering helio, architecture, monorepo structure, tech stack and commands.
dynamic-mcp AGENTS.md
Instructions for asyrjasalo/dynamic-mcp, covering dynamic-mcp knowledge base, structure, code map, where to look and conventions.
mcp-async copilot-instructions.md
Copilot instructions for mako10k/mcp-async, covering mcp async bridge server instructions, プロジェクト要件, 機能設計, 実装済み機能 and アーキテクチャ.
CC-Bridge AGENTS.md
AGENTS.md instructions for xhqing/CC-Bridge, a project described as: Claude Code upstream bridge framework (GLM/DeepSeek/MiMo/Kimi/Qwen) — routes Claude Code to third-party models via spoofed-whitelist IDs, per-model thinking levels configured in the bridge config, multi-key failover | Claude Code 上游桥接框架，通过白名单模型 ID…
dynamic-mcp CLAUDE.md
Instructions for asyrjasalo/dynamic-mcp, a project described as: MCP proxy server that reduces LLM context overhead with on-demand tool loading from multiple upstream servers.
openrouter-mcp-multimodal AGENTS.md
AGENTS.md instructions for stabgan/openrouter-mcp-multimodal, covering agent instructions, before you ship, releasing (read this before publishing), short version and version files (must all match package.json).