301-constraint-optimization

A set of speculative rules for choosing how strongly to restrict or tune a system, centered on testing about 60% as a starting point. The excerpt presents examples and hypotheses rather than a proven general method.

In plain words
What is it for?
It is for testing different resource, coverage, or process limits and looking for the setting that performs best. The provided material does not establish that 60% is reliably optimal.
Why use it?
It offers a starting experiment when the right constraint level is unknown, while encouraging comparisons across several settings.

Cursor rule

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/hamzaamjad/cursor-rules/301-constraint-optimization
Clone the repo
git clone --depth 1 https://github.com/hamzaamjad/cursor-rules
Per session 0 Nothing until a file matches its globs; then the whole rule loads.
When invoked 970 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 $0.00000 $0.00970
Opus 5 $0.00000 $0.00485
Sonnet 5 $0.00000 $0.00194
Haiku 4.5 $0.00000 $0.00097

Measured yesterday against content hash 1c5697049aba, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

301-constraint-optimization 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 yesterday.

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.

rules/300-techniques/301-constraint-optimization.mdc · 116 lines

How it starts

The opening of the file, as written. The whole thing — 116 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Constraint Optimization Patterns

Purpose

Guide optimization of system constraints based on the discovery that multiple AI agents independently converged on ~60% as the optimal constraint level across diverse domains. This rule provides heuristics for finding optimal constraints in new systems.

The 60% Principle

Multiple independent discoveries revealed that optimal performance often emerges around 60% constraint level:

  • Memory Allocation: 60% allocation for data processing systems
  • Test Coverage: 60% entropy/chaos in testing strategies
  • Technical Debt: Retaining 60% of "debt" maximizes velocity
  • Biological Rhythms: 60% of baseline (e.g., sleep) can optimize certain outputs

This suggests a universal principle: moderate constraints catalyze optimal performance.

Optimization Strategies

1. Starting Heuristic

When optimizing any system constraint without prior knowledge:

  • Begin testing at 60% as the initial hypothesis
  • Test range: 40%, 50%, 60%, 70%, 80%
  • Look for inverse U-curve performance patterns

2. Golden Ratio Consideration

The 60% principle (0.6) is remarkably close to the golden ratio (0.618):

  • Consider testing at exactly 61.8% for systems with aesthetic or natural components
  • Check Fibonacci ratios: 38.2%, 50%, 61.8%, 78.6%

3. Domain-Specific Variations

While 60% is common, expect slight variations:

  • Physical systems: May optimize at 55-65%
  • Cognitive tasks: Often exactly at 60%
  • Biological systems: Can range 50-70% depending on circadian factors
  • Creative work: Strong convergence at 60%

4. Testing Methodology

Use "superposition testing" when possible:

  • Test multiple constraint levels simultaneously
  • Run parallel experiments to reduce time
  • Look for resonance patterns across related systems

5. Fractal Properties

The principle shows self-similarity:

  • If 60% is optimal, test 36% (60% of 60%)
  • Look for harmonic relationships at multiples/fractions

Implementation Guidance

For Resource Allocation

# Example: Memory allocation
total_memory = system.available_memory()
optimal_allocation = total_memory * 0.6  # Start here
test_range = [0.4, 0.5, 0.6, 0.7, 0.8]  # Expand if needed

Read the full file on GitHub · 116 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. yesterday First seen · 116 lines · 0 tokens per session scan A 1c5697049aba

Subscribe to this mod's changes

301-constraint-optimization is a cursor rule published in the GitHub repository hamzaamjad/cursor-rules (2 stars, last pushed 1y ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 970 tokens. 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.