HerringtonDarkholme

Threads, server functions, GPU kernels, and distributed training all need rules for values used across an execution boundary. The shared language-design problem is to specify which values are admissible and how an accepted value becomes usable at the destination.

Working thesis: boundary crossing can be part of a language’s type and semantic contract. For a given boundary, the language can define admissible values, their destination types or representations, and the operations that copy, move, share, serialize, or proxy them. A compiler can check the applicable rules and arrange the corresponding runtime operations.

The central idea

The central proposal has two connected parts:

Define which values may be used across a boundary, and define what sending or sharing each accepted value means.

HerringtonDarkholme

CodeGraph Architecture — How the Code-Understanding Features Are Implemented

Audience: anyone (human or agent) who wants to understand how CodeGraph turns a codebase into a queryable knowledge graph and serves it to AI agents. Scope: a mechanism-level walkthrough of the whole pipeline, with file:line landmarks. For the single source of truth on agent-facing tool guidance see src/mcp/server-instructions.ts; for design history of specific mechanisms see docs/design/.

---

HerringtonDarkholme

Error in user YAML: (<unknown>): found character that cannot start any token while scanning for the next token at line 6 column 10

---
id: ai-lang-specification
title: AiLang - A Contextual Interface Definition Language for AI Agents
summary: AiLang is a declarative format that unifies tool interfaces, type systems, capability controls, domain knowledge, and usage examples into a single, AI-readable document. It enables agents to understand not just what tools exist, but when to use them, how to handle failures, and what contextual knowledge is required for effective operation. Features hierarchical detail levels for efficient context management.
loading: manual
keywords: specification, interface-definition, type-system, capability-model, ai-tools, runtime, tooling, context-management, detail-levels
related: @import(./manifesto.md), @import(./send_tweet.md)
version: 0.3.0
---

HerringtonDarkholme

Next.js Server Actions RCE Vulnerability Analysis

Executive Summary

This document analyzes a Remote Code Execution (RCE) vulnerability in React's Flight protocol, which powers Next.js Server Actions. The exploit chains three vulnerabilities:

1. Unsanitized path traversal in reference resolution (__proto__, constructor access)
2. Fake chunk injection - crafted objects treated as internal Chunk objects
3. Function constructor injection - _formData.get replaced with Function

HerringtonDarkholme

// a generic node for example
interface GNode { text: string; kind: string; next(): GNode | null; children(): GNode[]; }
// Process one node during traversal
type Process<T> = (n: GNode, t: Traversal) => Promise<T> | T
// continue the traversal to process children/siblings
interface Next { <T>(f: Process<T>): Promise<T> | T }
// traversal context
interface Traversal {
  // past nodes will be ready in Array
  ancestors: GNode[]; prevSiblings: GNode[]

HerringtonDarkholme

Q: tell me how Gemini family models can scale context to millions of token?

A:

The Gemini family of models achieves its impressive context window scaling through a combination of architectural innovations and advanced training techniques. Here's a breakdown of how they manage to handle millions of tokens:

1. Large Context Windows:

• Gemini 1.5 Pro: This model boasts a standard context window of 2 million tokens. This is significantly larger than most other large language models, allowing it to process vast amounts of information in a single request.
• Gemini 1.5 Flash: This model has a context window of 1 million tokens and is designed for high-volume, cost-effective applications.

HerringtonDarkholme

function findCombinations(numbers, targetSum) {
    const result = [];
    function find(currentSum, index, currentCombination) {
        if (currentSum === targetSum) {
            result.push([...currentCombination]);
            return;
        }
        if (currentSum > targetSum || index === numbers.length) return;

        // Include the number at the current index.

HerringtonDarkholme

React Effect Syntax Extension

This syntax extension is a mere fantasy of future React Script syntax. It is not a RFC, proposal or suggestion to React team or community. It is just a fun idea to imagine how React could be in the future.

Read this like a fan-fiction, for fun. :)

Background

There are two recent new features in React community that are introducing new extensions to JavaScript, both syntax and semantics:

HerringtonDarkholme

#define ｴｽﾃｨｰﾃﾞｨｰ std
#define ｱｲｵｰｽﾄﾘｰﾑ <iostream>
#define ﾕｰｼﾞﾝｸﾞ using
#define ｲﾌ if
#define ｲﾝｸﾙｰﾄﾞ #include
#define ｲﾝﾄ int
#define ｼｰｱｳﾄ cout
#define ｼｰｲﾝ cin
#define ﾈｰﾑｽﾍﾟｰｽ namespace
#define ﾌﾞｰﾙ bool

HerringtonDarkholme

const emailRef = ref('')

const userBio = server(async () => {
  const findUser = await import('./server')
  const userRef = ref(null)
  watch(async () => {
    userRef.value = await findUser(emailRef.value)
  })
  return (props) => (
    h('div', [userRef.value.bio]),