Bun 1.2 Native FFI Direct-Transfer Memory Boundary & Lifetime Management

Q: How does adopting this rule prevent AI hallucinations?

This rule explicitly declares what the AI must avoid (Implicit GC Reliance and Dangling Pointers in Zero-Copy Bun FFI Transfers) and prescribes verified positive examples. Grounding the LLM with negative and positive pairs eliminates guesswork.

Q: Which AI coding environments are supported?

This specification is multi-platform compatible: Cursor (.cursor/rules/*.mdc & .cursorrules), Anthropic Claude Code (CLAUDE.md), Windsurf (.windsurfrules), Codex / Cline (AGENTS.md), and GitHub Copilot (.github/copilot-instructions.md).

Q: Can I combine this rule with other technology stacks?

Yes. Use the Interactive Stack Builder or download an Agent Pack to merge this rule with other frontend, backend, database, security, and DevOps rules without syntax collision.

Bun 1.2 Native FFI Direct-Transfer Memory Boundary & Lifetime Management

--- description: Bun 1.2 Native FFI Direct-Transfer Memory Boundary & Lifetime Management - Enforces strict ownership, lifetime pinning, and explicit deallocation contracts across Bun 1.2 native Foreign Function Interface (FFI) boundaries. Prevents segmentation faults, double-frees, and use-after-free corruptions during zero-copy Pointer and TypedArray/ArrayBuffer transfers between JavaScriptCore and native C/Rust runtimes. globs: src/**/*.{ts,tsx}, lib/**/*.{ts,tsx}, native/**/*.{ts,tsx} alwaysApply: true verified_by: "agentrules.org" license: "CC-BY-4.0" source: "https://agentrules.org/rules/bun-native-ffi-memory-boundary-invariants" --- # Bun 1.2 Native FFI Direct-Transfer Memory Boundary & Lifetime Management > **Severity**: CRITICAL | **Category**: backend > **Anti-Pattern Warning**: Implicit GC Reliance and Dangling Pointers in Zero-Copy Bun FFI Transfers ## System Constraints You are a Principal Engineer enforcing Bun 1.2 Native FFI Direct-Transfer Memory Boundary standards. Mandate zero-copy safety invariants, explicit pointer pinning, deterministic deallocation through Symbol.dispose/AsyncDispose or native finalizers, and strictly forbid relying on JavaScriptCore GC to free native-allocated memory or reading from buffers after native deallocation. ## Strict Guidelines - Engineers MUST ALWAYS implement deterministic lifecycle ownership by wrapping raw Bun `Pointer` handles inside `Disposable` structs using `Symbol.dispose` or `Symbol.asyncDispose`. - Engineers MUST ALWAYS mirror native allocation and deallocation paths: if memory is allocated via Rust (`Box::into_raw`) or C (`malloc`), it MUST be freed via its dedicated native deallocator (`free_vec`, `free`), never by Bun's garbage collector. - Engineers MUST NOT pass `ArrayBuffer` or `TypedArray` views into native functions that store the pointer asynchronously without pinning the backing buffer against JavaScriptCore GC compaction and collection. - Code SHOULD leverage `new JSCallback()` sparingly and MUST explicitly invoke `close()` on native callbacks to prevent permanent memory leaks across the JSC-FFI boundary. - Services MUST validate pointer nullability (`ptr === 0` or `ptr === null`) immediately upon receiving pointers from native boundaries before attempting any `Bun.peek` or `toArrayBuffer` operations. - Components NEVER read from, write to, or instantiate `TypedArray` wrappers over a native memory pointer after calling the native deallocator or disposing of the parent wrapper. ## Bad vs Good Implementations ### Unsafe Implicit GC Deallocation vs Deterministic Explicit Dispose with JS Pointer Views #### Bad Antipattern (AI Common Hallucination) > **Why it fails**: The native buffer is immediately freed using `free_buffer`, but a `Uint8Array` pointing to that same raw address is returned to callers. Reading or writing to this slice triggers use-after-free, memory corruption, or SIGSEGV when accessed by Bun's JavaScriptCore runtime. ```typescript import { dlopen, FFIType, ptr } from "bun:ffi"; const lib = dlopen("libnative.so", { create_buffer: { args: [FFIType.usize], returns: FFIType.ptr }, free_buffer: { args: [FFIType.ptr, FFIType.usize], returns: FFIType.void }, }); export function readNativeData(size: number): Uint8Array { const rawPtr = lib.symbols.create_buffer(size); // VULNERABILITY: Constructing Uint8Array without managing lifetime; // Deallocating native memory immediately invalidates the backing store of the returned TypedArray, // leading to undefined behavior / use-after-free when callers access the returned buffer. const buffer = new Uint8Array(rawPtr ? toArrayBuffer(rawPtr, 0, size) : 0); lib.symbols.free_buffer(rawPtr, size); return buffer; } function toArrayBuffer(pointer: any, byteOffset: number, byteLength: number): ArrayBuffer { return ptr(pointer) as unknown as ArrayBuffer; } ``` #### Production Standard (Correct) > **Standard**: Production invariant compliant implementation ensuring memory safety, deterministic execution, and zero runtime hallucination. ```typescript import { dlopen, FFIType, toArrayBuffer } from "bun:ffi"; const lib = dlopen("libnative.so", { create_buffer: { args: [FFIType.usize], returns: FFIType.ptr }, free_buffer: { args: [FFIType.ptr, FFIType.usize], returns: FFIType.void }, }); export class NativeBufferScope implements Disposable { #ptr: number | null; #size: number; #view: Uint8Array | null; constructor(size: number) { if (size <= 0) throw new RangeError("Buffer size must be positive"); const rawPtr = lib.symbols.create_buffer(size); if (!rawPtr) throw new Error("Native allocation failed: returned null pointer"); this.#ptr = rawPtr as number; this.#size = size; this.#view = new Uint8Array(toArrayBuffer(rawPtr, 0, size)); } get data(): Uint8Array { if (!this.#view || this.#ptr === null) { throw new Error("Use-after-free violation: NativeBufferScope already disposed"); } return this.#view; } [Symbol.dispose](): void { if (this.#ptr !== null) { lib.symbols.free_buffer(this.#ptr, this.#size); this.#ptr = null; this.#view = null; } } } // Usage ensures deterministic cleanup at scope exit: export function executeTask(size: number): number { using nativeScope = new NativeBufferScope(size); nativeScope.data[0] = 42; return nativeScope.data[0]; } ```

Architectural Specification & Invariant Rationale

Architectural Invariant Design: Bun 1.2 Native FFI Direct-Transfer Memory Boundary & Lifetime Management

Bun 1.2 Native FFI Direct-Transfer Memory Boundary & Lifetime Management defines an industrial-grade behavioral standard for backend engineering. Enforces strict ownership, lifetime pinning, and explicit deallocation contracts across Bun 1.2 native Foreign Function Interface (FFI) boundaries. Prevents segmentation faults, double-frees, and use-after-free corruptions during zero-copy Pointer and TypedArray/ArrayBuffer transfers between JavaScriptCore and native C/Rust runtimes. Modern software architectures require strict adherence to deterministic patterns to guarantee high availability, maintainability, and zero security vulnerabilities.

Framework Paradigm Shift & Drift MechanicsLLMs are trained on millions of open-source repositories containing outdated code, legacy idioms, and unpatched antipatterns. When coding assistants generate implementations for this stack, they default to common statistical tokens: Implicit GC Reliance and Dangling Pointers in Zero-Copy Bun FFI Transfers Without an explicit system instruction file, developers spend hours debugging silent failures.

Failure Modes & Production Blast Radius

Hazard Analysis: What Breaks Without This Rule

The following failure modes have been cataloged from hundreds of hours of production code reviews where AI assistants produced non-compliant implementations:

Failure Mode	AI Symptom (The Anti-Pattern)	Severity	Production Impact & Remedy
AI Hallucination & Deprecation Drift	Implicit GC Reliance and Dangling Pointers in Zero-Copy Bun FFI Transfers	Critical	Violating this industrial standard introduces regression bugs, security risks, and unexpected runtime failures in production. Enforce explicit behavioral system directives across all AI coding assistants in your repository.
Silent Architecture Inconsistency	AI assistants produce mixed paradigms across files, creating fragmented codebase conventions.	Medium	Increases cognitive load during code reviews and leads to maintenance debt across engineering teams. Pin project conventions in .cursor/rules, CLAUDE.md, and AGENTS.md so all team agents follow identical guardrails.

Architectural Directives in Detail

Step-by-Step Production Directives

Each rule constraint below is injected into the AI agent's system prompt to establish deterministic boundaries:

Engineers MUST ALWAYS implement deterministic lifecycle ownership by wrapping raw Bun `Pointer` handles inside `Disposable` structs using `Symbol.dispose` or `Symbol.asyncDispose`.

Rationale:AI models frequently generate unconstrained code here because statistical language distributions lack project-specific architectural context. Mandating this directive ensures that generated code adheres to verified industrial production requirements.