
Mirage In-Browser WASM Execution Sandbox & Web Worker Jail
A zero-latency, client-side WebAssembly execution sandbox for DeepSeek Harness. Runs Python, JavaScript, and SQLite inside isolated Web Workers without backend server overhead.
WASM-Sandbox · In-Browser · Pyodide · Client-Side
$ pnpm add @strukto-ai/mirageArchitecture
Core Features
Pure Client-Side WASM Execution: Zero backend container hosting costs or cluster maintenance
Airgapped Data Privacy: Sensitive datasets process locally within the user's browser securely
Sub-Millisecond Cold Starts: Eliminates multi-second container boot lags for instantaneous feedback
Scientific Python Ecosystem: Full support for NumPy, Pandas, Matplotlib, and SymPy natively
Core Workflow
Web Worker Jail Spawning
Spawns an isolated Web Worker thread with an in-memory virtual filesystem.
WASM Runtime Hydration
Hydrates Pyodide or QuickJS WASM micro-kernels and standard library modules.
In-Memory Code Evaluation
Executes target scripts with zero access to main-thread DOM or session cookies.
Output & Canvas Pipe
Pipes stdout, return values, and synthesized chart images back to the chat UI.
Configuration Parameters Reference (YAML / JSON)
| Parameter | Type | Default | Description |
|---|---|---|---|
| runtimeLanguage | string | python | WASM sandbox runtime engine |
| executionTimeoutMs | number | 10000 | Worker execution timeout in milliseconds |
| preloadStandardLibs | boolean | true | Pre-warm scientific libraries in background |
Recommended Plugins
dsh-purge Red-Team Eval & Model Plugin
dsh-purge is an open-source cybersecurity red-team evaluation plugin and harness jailbreak toolkit designed for DeepSeek Harness. Decoupling host-level runtime guardrails from intrinsic model responses, it enables custom system prompt overrides, dynamic rule set switching, and local security testbeds. Operating strictly within authorized local environments, it empowers developers to assess large language model robustness.
Gungnir Hardened Container Sandbox & Isolation Jail
A zero-trust hardened container sandbox for executing untrusted AI code. Powered by Linux namespaces, strict seccomp syscall filters, and read-only rootfs with sub-millisecond cold starts.
Discipline Guard Policy Interceptor & Safety Filter
A bidirectional safety firewall and policy enforcement proxy for DeepSeek Harness. Prevents prompt injection, redacts sensitive PII data, and intercepts dangerous CLI operations in real time.