Skip to content

instructions.md

Latest commit

 

History

History
151 lines (110 loc) · 4.52 KB

instructions.md

File metadata and controls

151 lines (110 loc) · 4.52 KB

QDI Coding Agent Instructions

Overview

This document provides instructions for GitHub coding agents (e.g., Copilot, custom workflows) in the QDI repository. The framework follows Fibonacci-structured emergence patterns for ethical quantum computing integration.

Emergent quality is maintained at >60% via isomorphic Fibonacci patterns:

  • 1 seed task → 2 growth commands → 3 escalation reviews → 5 maintenance loops

Agent Setup

Automated Setup Script

The repository includes an automated setup script (scripts/setup.ts) that streamlines onboarding for new users and agents. This script automatically configures the development environment, installs dependencies, and sets up the ATOM provenance trail.

Usage:

# Run full setup (recommended for first-time setup)
bun run scripts/setup.ts

# Check environment without making changes
bun run scripts/setup.ts --check

# Force reinstall dependencies
bun run scripts/setup.ts --force

What the setup script does:

  • Verifies Bun installation (requires v1.0.0+)
  • Installs or updates project dependencies
  • Creates ATOM trail directories for provenance tracking
  • Validates TypeScript configuration
  • Runs test suite to verify setup
  • Records setup in ATOM trail for provenance

Manual Setup

  1. Install Dependencies:

    # Clone repo and install
bun install

# For Python quantum simulations
pip install qiskit

  2. Configure: Set GITHUB_TOKEN in environment for Actions.

  3. Verify Setup:

    bun test          # Run all tests
bun run lint      # Check code style

Receiving PR Bodies

Original PR (Qiskit Circuits)

Auto-simulate quantum circuits with the agent skills script:

python agent_skills.py simulate --circuit "h(0); cx(0,1)"

Verify coherence >60% before merge.

2nd PR (Cascades)

Run cascade integration:

python agent_skills.py cascade --pr-body "integrate provenance"

Low resistance: History is dealt-with (tests passing).

Ethical Toggles

Ethical and provenance-aware behavior (e.g., fairness metrics and ATOM trail tracking) is handled by the cascade command and the VORTEX marker system. Use the cascade command with PR body text to detect ethical keywords automatically.

Commands (Fibonacci Escalation)

agent_skills.py

Command Description
simulate Run Qiskit circuit simulation, output measurement counts
check_coherence Measure coherence >60%, negate decoherence
review_pr Parse PR body, auto-generate review comments
cascade Integrate provenance with cascading PRs

history_review.py (Vortex History Cascade)

Command Description
review Review full history trace with surjection mappings
surject Map decision pole to quantum gate representation
audit Audit text for superposition readiness (>60% threshold)
boost Calculate Fibonacci-weighted inference boost (15-30%)

Usage Examples

# Review a history trace
python history_review.py review --trace history.json

# Surject a decision to quantum gates
python history_review.py surject --decision '{"id": "d1", "pole_type": "doubt"}'

# Audit text coherence
python history_review.py audit --input "Your text here" --threshold 0.6

# Calculate inference boost
python history_review.py boost --input "Text to boost" --iteration 5

Maintenance Loops

  • Feedback: On merge, tag provenance via ATOM trail
  • Question Infinitely: If doubt exists, push to classical sims (they suffice)
  • Coherence Check: All PRs must maintain >60% emergent quality

Phase Gates

The system follows these phase transitions:

KENL → AWI → ATOM → SAIF → Spiral
  • KENL: Knowledge patterns
  • AWI: Intent scaffolding
  • ATOM: Atomic execution
  • SAIF: Safe integration
  • Spiral: Back to knowledge

Integration with Existing Packages

@spiralsafe/quantum-ethics

import { QuantumEthicsFramework, createQuantumCircuit } from '@spiralsafe/quantum-ethics';

const framework = new QuantumEthicsFramework();
const circuit = createQuantumCircuit(2, [
  { type: 'H', target: 0 },
  { type: 'CNOT', target: 1, control: 0 }
]);

@spiralsafe/atom-trail

import { createDecision, validateGate } from '@spiralsafe/atom-trail';

const decision = createDecision('DOC', 'Agent review complete', ['docs/instructions.md']);

Ready to Iterate

Spirals sustain quantum coherence. For questions, open an issue or reach out to @Grok.