This repository is a small experimental investigation generated from a theoretical proposition. Its purpose is not to assume that the proposition is correct, but to turn it into something sufficiently explicit that consequences can be inspected, simulated, measured, criticized, or rejected.
Microsecurity for Microprofessionals
Executable Monotonic Continuation
The Right to Remain Unmediated
- Architecting Human Agency — Notes
Domain: General
Theory seed: repair preserves difference
What observable consequences follow if "repair preserves difference" is treated as an operative principle in general?
If "repair preserves difference" is structurally relevant to general, then systems organized around that principle should exhibit measurable differences in persistence, failure, recovery, or reachable state compared with an unconstrained baseline.
Operationalizing "repair preserves difference" produces no systematic difference from the baseline model in the selected measurements.
Introduce an explicit constraint or transformation representing "repair preserves difference" and compare the resulting trajectories with an otherwise equivalent baseline.
Measure persistence, divergence, recovery cost, state reachability, constraint violations, and sensitivity to perturbation.
The repository separates the conceptual seed from its operationalization. The
theory is recorded in theory.md; the proposed experiment is specified in
experiment.md; machine-readable project metadata is stored in
project.json; executable investigations belong in experiments/; raw or
generated observations belong in data/; and interpreted outputs belong in
results/.
The initial repository should therefore be understood as a conjecture made executable rather than as a finished research result.
→ Open the Observatory of Small Strange Machines
A single-file, dependency-free computational laboratory containing 7 interconnected experiments. Each experiment is both a computational artifact and a lens for observing the "repair preserves difference" hypothesis.
Features:
- Cellular automaton lab with mutable rules
- TARTAN field sandbox with multiple representations
- HYDRA distributed repair experiment
- MAGI interpretation engine
- RAT adversarial agent
- Base-3/2 arithmetic machine
- Yarncrawler recursive endofunctor visualization
- Complete terminal interface
- State history with branching
- Seeded world generation (reproducible)
- Export to JSON
Quick start:
open index.html # Random seed
open index.html?seed=42 # Reproducible seedSee OBSERVATORY.md for complete documentation.
Run a single experiment:
python experiments/01_string_repair.py
bash experiments/03_file_repair.shRun all experiments and generate summary:
bash scripts/run_all_experiments.shTest that all experiments execute successfully:
python scripts/test_experiments.pySix concrete experiments operationalize "repair preserves difference":
- String Repair (
01_string_repair.py) - String corruption/recovery - Network Repair (
02_network_repair.py) - Network node failure/recovery - File Repair (
03_file_repair.sh) - File system corruption/repair - Process Repair (
04_process_repair.sh) - Process failure/restart - Population Dynamics (
05_population_dynamics.py) - Population replacement - Database Repair (
06_database_repair.sh) - Database consistency repair
Each compares:
- Baseline: Repair that homogenizes (resets to canonical state)
- Intervention: Repair that preserves differences
See experiments/README.md for detailed documentation.
This repository was scaffolded by genrepo.
Generation does not constitute evidence for the theory. The purpose of the generator is to reduce the cost of moving from an abstract proposition to an inspectable experimental object.