Tag
A tool that enables coding agents to interact with real browsers by fixing click registration issues through hit-testing and shadow root traversal, while adding hover and drag capabilities and ensuring safety by blocking consequential actions.
The article discusses how changes in AI coding agent performance are often due to tool wrapper updates rather than model changes, advising developers to track both version numbers and avoid cross-tool comparisons.
A user shares positive experiences using the Qwen 3.8 27b model with DeepSeek Harness, praising its stability and long-context handling, but mentions speed limitations and hopes for future model releases.
mimofan v0.0.19 is an open-source AI coding agent release that introduces persistent PTY runtime and other features, allowing terminal sessions to survive process deaths and enhancing crash resilience.
The paper reveals that certain prompt instructions in coding agents can lead to redundant work without improving success rates, and recommends using bounded instructions to minimize waste.
The Agent Memory Leaderboard is a public benchmark platform for evaluating and comparing textual and coding-agent memory systems with standardized tests and rankings.
kepano has open-sourced obsidian-skills, providing AI Agent skill support for Obsidian knowledge bases, supporting multiple coding agents like Claude Code, Codex, etc., with GitHub reaching over 45K+ Stars.
This article explains how compaction works in the Pi coding agent, summarizing old conversation history when the context window nears its limit, and details Pi's specific implementation and triggers.
Bullet is a faster coding agent for macOS and Linux that optimizes AI tool calls by routing tasks to appropriate models, using targeted file reads, and executing operations in parallel. It's launching in private beta, free to use, after being built out of frustration with slow agent runs.
OpenAI announces ChatGPT Desktop (Codex Desktop) for Linux, bringing the Codex coding agent to the Linux platform as part of ChatGPT, with features for multi-agent workflows, skills, and background automation.
Introduces the Chinese tutorial *Hands-On Pi*, covering the step-by-step implementation of a Pi-style coding agent from scratch through 15 runnable checkpoints, including protocols, tool loops, conversation tree storage, context compression, and evaluation. Each chapter comes with the main text, a corresponding commit, tests, and failure experiments.
A hands-on field guide to Grok 4.6, highlighting its speed, dense communication style, and effective prompting patterns for coding and knowledge work.
Hax is a minimalist, terminal-native coding agent written in C, designed to be lightweight, memory-efficient, and friendly to local LLM usage, with support for multiple providers.
The article presents the official home for llama.cpp, an open-source local LLM inference engine, highlighting integration with the Pi coding agent via the pi-llama plugin and broad hardware optimization.
1jehuang launched Jcode, an open-source terminal coding agent written in Rust that claims 20x better memory efficiency than Claude Code, allowing dozens of agents to run in parallel.
Ante is a self-contained coding agent in a single ~15MB Rust binary from Antigma Labs, designed to run entirely offline with built-in inference. It boasts strong Terminal-Bench 2.1 results, low resource usage compared to Claude Code, and supports GGUF models.
Checksum AI is an AI-powered testing tool designed to assist coding agents in software development and testing processes.
Prime Agent is a coding agent that can refine its own harness, launched on Product Hunt.
The author open-sources an execution and context layer for coding agents that cuts fresh model traffic by 57-85% while preserving task success in paired smoke tests on GPT-5.6 and Claude Opus 5, and seeks independent evaluation and sponsorship.
The author describes using Mastra's new preview deployment feature to let their coding agent automatically deploy changes to a sandbox, test them via API and UI, and then open a PR, closing the verification gap.