opus-4-7-features

Optional instructions for using Claude Opus 4.7, an AI model, with its current reasoning and effort settings.

In plain words
What is it for?
It is for configuring adaptive thinking, selecting effort levels, guiding response depth in prompts, and using prompt caching or side questions where supported.
Why use it?
It helps avoid invalid settings and explains how to choose between deeper reasoning and faster responses when using that model.

Cursor rule for Cursor

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.

agentmods
npx agentmods add rules/gosha70/code-copilot-team/opus-4-7-features
Clone the repo
git clone --depth 1 https://github.com/gosha70/code-copilot-team

Made for: Cursor.

Per session 29 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,576 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00029 $0.01576
Opus 5 $0.00015 $0.00788
Sonnet 5 $0.00006 $0.00315
Haiku 4.5 $0.00003 $0.00158

Measured 2d ago against content hash d0dc537318a1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

opus-4-7-features 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 2d 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.

adapters/cursor/.cursor/rules/opus-4-7-features.mdc · 142 lines

How it starts

The opening of the file, as written. The whole thing — 142 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Opus 4.7 Features & Optimization

Optional guidance for sessions using Claude Opus 4.7 (v2.1.111+). Load this rule when you want to leverage Opus 4.7-specific capabilities.

Adaptive Thinking (Breaking Change)

Opus 4.7 changes the API contract for extended thinking. The legacy form is rejected:

  • thinking: {type: "enabled", budget_tokens: N} → returns 400 Bad Request on Opus 4.7.
  • Only thinking: {type: "adaptive"} is accepted.

With adaptive thinking, the model dynamically decides when and how much to think based on query complexity. There is no fixed token budget — Claude self-regulates depth.

How to control depth:

  • Use the effort parameter (xhigh / high / medium / low) instead of budget_tokens.
  • Thinking is also promptable in-message:
    • "Think carefully and step-by-step" → encourages deeper reasoning.
    • "Prioritize responding quickly" → encourages a shorter thinking pass.

Older models: enabled/budget_tokens is deprecated on Opus 4.6 and Sonnet 4.6 (still functional today, but on the way out). Migrate to adaptive proactively so your code keeps working when the deprecation lands.

Effort Level: xhigh

Opus 4.7 introduces a new effort level xhigh between high and max. Other models fall back to high when xhigh is requested.

Effort When to use Phase fit
low Quick lookups, file searches, trivial edits
medium Standard implementation, build tasks Build
high Architecture, planning, detailed review Research, Plan, Review
xhigh Complex multi-file reasoning, deep architectural analysis, hard debugging Plan (large features), Review (cross-cutting PRs)

Availability: xhigh became selectable in Claude Code v2.1.111+ when running Opus 4.7.

Default change: From Claude Code v2.1.117+, xhigh is the default effort level for Opus 4.7. On v2.1.111–v2.1.116 the default is still high and you must opt in with /effort xhigh. Other models continue to default to high regardless of Claude Code version.

Read the full file on GitHub · 142 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. 2d ago First seen · 142 lines · 29 tokens per session scan A d0dc537318a1

Subscribe to this mod's changes

opus-4-7-features is a cursor rule published in the GitHub repository gosha70/code-copilot-team (6 stars, last pushed 2d ago), licensed MIT. It adds 29 tokens to every session and 1,576 once invoked, about $0.0001 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.