minimax-m3-long-context

minimax-m3-long-context is a skill for Cursor from madebyaris/advance-minimax-m3-cursor-rules. It costs 101 tokens per session (1,494 once invoked), scanned A, original, MIT.

A guide for using MiniMax M3's 1-million-token context window, including what to load fully, what to summarize, and what to retrieve later.

In plain words
What is it for?
Use it for large codebases, long transcripts, multi-file refactors, or investigations expected to require several rounds of work.
Why use it?
It helps prevent very large projects or research tasks from filling the context with information that does not need to remain verbatim.

Skill for Cursor

Written for Cursor: installed under .cursor/.

Good fit Use it for large codebases, long transcripts, multi-file refactors, or investigations expected to require several rounds of work.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/madebyaris/advance-minimax-m3-cursor-rules/minimax-m3-long-context
Install

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.

Any agent
npx skills add madebyaris/advance-minimax-m3-cursor-rules --skill minimax-m3-long-context
Clone the repo
git clone --depth 1 https://github.com/madebyaris/advance-minimax-m3-cursor-rules

Made for: Cursor.

Wrote 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.

agentmods badge for minimax-m3-long-context

README.md
[![agentmods](https://agentmods.dev/badge/skills/madebyaris/advance-minimax-m3-cursor-rules/minimax-m3-long-context/github.svg)](https://agentmods.dev/skills/madebyaris/advance-minimax-m3-cursor-rules/minimax-m3-long-context)
Your own site
<a href="https://agentmods.dev/skills/madebyaris/advance-minimax-m3-cursor-rules/minimax-m3-long-context"><img src="https://agentmods.dev/badge/skills/madebyaris/advance-minimax-m3-cursor-rules/minimax-m3-long-context/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for minimax-m3-long-context

Your own site · 80×15
<a href="https://agentmods.dev/skills/madebyaris/advance-minimax-m3-cursor-rules/minimax-m3-long-context"><img src="https://agentmods.dev/badge/skills/madebyaris/advance-minimax-m3-cursor-rules/minimax-m3-long-context.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 101 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,494 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00101 $0.01494
Opus 5 $0.00051 $0.00747
Sonnet 5 $0.00020 $0.00299
Haiku 4.5 $0.00010 $0.00149

Measured 13d ago against content hash 251fd2ca8cec, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

minimax-m3-long-context 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 13d 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.

.cursor/skills/minimax-m3-long-context/SKILL.md · 128 lines

How it starts

The opening of the file, as written. The whole thing — 128 lines — stays where its author put it; the contents beside it link to each section on GitHub.

M3 Long-Context Discipline

M3 ships a 1M-token MSA context window. The room is large; the cost of using it badly is also real. This skill teaches the retention and compression decisions that keep long-context work honest.

When to Use

  • The full content (files, search results, fetched pages, transcripts, design notes) might exceed ~200K tokens.
  • The user explicitly asks to "keep all of this in mind", "use the whole repo", or "don't lose anything".
  • You are tempted to start a fresh session to "free context" — that is usually a compression failure, not a context failure.
  • Multi-file refactors across a large codebase, transcript analysis, full-repo synthesis, or retrieval-augmented synthesis.
  • A research / debugging / migration task that you expect to iterate more than 3 times.

For a single-file edit or a small bug fix, you do not need this skill.

Step 0: Decide Retention Per Slice

For each chunk of evidence you are about to load, pick one of three retention modes before you load it:

  • Keep verbatim — the file is the answer, the user asked to see it, or the next step depends on exact contents.
  • Keep summary — the contents matter for context but you only need the high-signal lines.
  • Drop — the chunk is tangential, redundant with something already in context, or only useful for one specific iteration that has passed.

This is the same as deep-research Phase 2's "drop tangential" rule, applied at the file level before loading.

Step 1: Plan The Loader

Before the first read or search, write a 4–6 line plan in your scratchpad:

Loader plan
  In context at start: [system + always-on rules + user task]
  Add verbatim:      [the few files the answer depends on]
  Add as summary:    [reference docs, fetched pages, prior search results]
  Drop:              [tangential files, duplicate docs, raw search output past its iteration]
  Compress at:       [end of each iteration; before any new search round]

If you cannot write this plan, the task is under-specified — go back to the user or the codebase.

Read the full file on GitHub · 128 lines

Changes

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.

  1. 13d ago First seen · 128 lines · 0 tokens per session scan A 251fd2ca8cec

Subscribe to this mod's changes

minimax-m3-long-context is a skill published in the GitHub repository madebyaris/advance-minimax-m3-cursor-rules (125 stars, last pushed 2mo ago), licensed MIT. It adds 101 tokens to every session and 1,494 once invoked, about $0.0005 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 2026-08-30.