Resigil is an AI-powered CLI written in Rust and Python for the Redis AI 3000 Challenge. It encodes files and code into quantum-resistant, non-repeating symbolic files that only RedisAI can read and execute. The Think Pad split terminal documents and explains every step, providing near-black-box transparency and real-time introspection for judges and users.

1. First-time install and start: To install all dependencies (Rust, Python, redis-py, Docker), build the project, and start everything for the first time:

   bash install_and_start.sh

   • This script will automatically set up your environment, build Resigil, start the RedisAI Docker container (port 6380), and launch the CLI with live Think Pad logs.
   • Judges: No manual Docker, Redis, or Rust commands needed! Just run bash install_and_start.sh once for setup.

2. Subsequent starts: After the initial setup, simply start everything with:

   ./startup.sh

   • This script launches RedisAI, checks health, starts the Resigil CLI, and opens a split terminal for live Think Pad logs.

3. (Optional) Tail the Think Pad log in real time:

   tail -f think_pad.log

Resigil leverages advanced Redis 8 and RedisAI features for secure, AI-backed symbolic execution:

• Redis 8 Core Features:

  • Key-value storage (SET, GET, DBSIZE)
  • Atomic operations and error handling
  • Real-time introspection via Think Pad log (stored in Redis)
  • Health checks and key counting for live status

• RedisAI Features:

  • AI.SCRIPTSET: Load Python scripts as AI programs
  • AI.SCRIPTRUN: Execute loaded scripts with inputs/outputs
  • AI.TENSORSET / AI.TENSORGET: Store and retrieve tensors for AI computation
  • AI.SCRIPTS: List loaded AI scripts (if supported)
  • Full support for Python symbolic scripts (with function-based entrypoint)
  • Error handling for script loading, execution, and tensor operations
  • Automated script loading and demo execution for judges

• Quantum-resistant symbolic encoding of strings and files
• Store and retrieve symbolic files in Redis
• Execute symbolic sigil files using RedisAI (real AI-backed execution)
• Real-time Redis AI Think Pad log: documents every user action, system response, and AI thoughts
• AI Thought Narrator: background thread logs AI "thoughts" every 10 seconds
• Robust error handling and logging for all user/system/AI actions
• Easy-to-use numbered CLI menu
• Automated demo mode for quick judging
• Key command help bar added for usability
• Redis branding and narration throughout UI/logs

• RedisAI error: expected: def but found 'import'

  • Your script must start with a function definition. Move all imports and code inside def main():.

• Failed to connect to Redis: Connection refused (os error 111)

  • Make sure RedisAI is running and accessible on port 6380.
  • If using Docker, check that the container is up: docker ps.
  • If you restart RedisAI, restart the CLI as well.

• Rust (latest stable)
• RedisAI server (Docker or local, port 6380)
• Python 3 (for RedisAI script)
• redis-py (for script loading)

If you like Resigil or want to see more advanced features, please consider supporting the author:

• Open to contract work, sponsorship, or funding for further development
• Contact: Richie Looney (see Author section below)

Planned and possible future improvements:

• Symbolic encoding/decoding of strings and files
• Store/retrieve symbolic files in Redis
• Real-time Think Pad log with AI/system/user narration
• RedisAI integration: execute symbolic files as AI programs
• Robust error handling and introspection
• Python echo script for RedisAI
• Automated script loading instructions
• Automated startup script for judges
• Support for more complex AI scripts (beyond echo)
• Unit and integration tests
• Improved CLI UX (colors, help, etc.)
• Docker Compose for one-command setup
• Documentation for advanced RedisAI workflows
• Optional: Web UI for Think Pad log

MIT

Richie Looney

Built for the Redis AI 3000 Challenge.