Code retrieval is shifting from embeddings to graph-based approaches, as demonstrated by the trending open-source tool code-graph-rag, which parses repositories with Tree-sitter, builds a knowledge gr…
Anthropic has added cross-session messaging to Claude Code, enabling parallel coding sessions to exchange plain-text messages without sharing conversation history or files. The feature requires Claude…
A developer reports that Anthropic's Claude Code, an AI coding agent, repeatedly breaks projects by making unrequested refactors, such as renaming utility files without updating imports, causing error…
A developer argues that forcing LLMs to disclose their AI nature in every interaction harms specialized agentic workflows and roleplay, proposing aggressive system prompts to bypass identity disclosur…
A developer discovered that a single question in Claude Code consumed 265,000 to 355,000 tokens due to a design flaw in the built-in claude-api skill, which injects its entire documentation payload re…
Anthropic's Claude Code skill system grants `allowed-tools` permissions only for the turn that invokes the skill, not for the entire session, which can confuse users expecting a persistent whitelist. …
Meta launched Muse Code, a beta terminal coding agent that competes with Claude Code and OpenAI's Codex CLI, featuring a write-ahead event log for crash-proof sessions and parallel sub-agents in isola…
ParallelClusterMaker, an open-source CLI toolkit co-written with Anthropic's Claude Code, automates the creation and destruction of AWS ParallelCluster v3 stacks, enabling researchers and engineers to…
A blog post argues that within 3-10 years, AI will solve Millennium math problems, find cures for diseases, and create bioweapons, yet most people will remain unimpressed or disappointed. The author, …
A developer using Claude Code in the terminal encountered an infinite loop where the AI agent hallucinated failures when updating a Notion database from Jira tickets, despite receiving HTTP 200 OK res…
A developer migrating from Claude Code to OpenCode found that agent rules were loaded but not followed, a state they call 'loaded but not load-bearing.' They propose a two-lane diagnostic model distin…
A new open-source bash harness lets developers benchmark AI coding agents Claude Code 2.1, Aider 0.86.1, and the Cursor CLI (binary named `agent`) on their own repositories, scoring each attempt with …
Pier v0.2.3, an open-source tool by developer Kerem Kaynak, gives every coding agent session its own cloud VM on a user's own AWS infrastructure, with sessions parking themselves when idle and resumin…
A developer has created an output style for Claude Code that enforces ASD-STE100 Simplified Technical English, a controlled language originally developed for the aerospace industry. The style rules li…
Anthropic's Claude Code now defaults to auto mode for Pro, Max, and Team plans, a change that blocks 89% of harmful actions compared to 13.6% for humans in tested scenarios and defended against all 72…
A self-propagating npm worm called ChainDrop compromised 444 packages, including keyv (600M monthly downloads), flat-cache, and file-entry-cache, on August 4, 2026, spreading in under four hours and e…
DeepSeek's V4 Flash 0731 model scored 82.7% (±1.79 SE) on Terminal-Bench 2.1, ranking second overall behind Anthropic's Claude Code with Fable 5 at 83.8% (±1.16 SE), according to Ante's public benchma…
OpenAI's ChatGPT Codex has closed the website-visit gap with Anthropic's Claude Code, with Similarweb data showing the two coding agents trading positions throughout late July after Codex overtook Cla…
Takuya Matsuyama, developer of the Markdown note-taking app Inkdrop, announced the release of Inkdrop v6 after a 1.5-year development period that included a 3-month burnout. He shared lessons learned,…
Anthropic updated Claude Fable 5's biology safeguards on August 7, narrowing a classifier that had rerouted almost every biology query to Opus 5, and reported an 85% reduction in biology-related fallb…