Anthropic's February 2026 study classified 11 collaboration behaviors across 9,830 Claude conversations, forming the AI Fluency Index. A developer built skill-tree, a tool that analyzes Claude Code or…
A developer built an unattended autopilot that uses Claude Code to improve its own environment daily, with cost and repetition safeguards. The system runs at 5 AM via launchd, picking one improvement …
Sanbox, a new platform for running AI agents in isolated and resumable sandboxes, has been launched. The platform uses MicroVM isolation, a persistent filesystem, and a live trail of run events, with …
A developer created PerformanceAgent, an open-source AI strength and conditioning coach that runs inside AI agent CLIs like Claude Code. The tool uses a deterministic Python engine for calculations an…
Spendict launches a deterministic AI tool that evaluates advertisements before any budget is spent, providing a run, fix_first, or kill verdict based on performance marketing expertise. The tool integ…
Anthropic analyzed 1.2 million Claude Cowork sessions and found that about half of all usage goes toward mundane office tasks like compiling status reports and building onboarding checklists, rather t…
LiteLLM's pricing table, used by most AI cost tools, has only a single automated test—`jq empty`—which verifies JSON parsing but not value accuracy. An audit found records citing their own refutation,…
Developer dotbrt released DejaView, an open-source terminal dashboard that aggregates and displays all Claude Code sessions across projects on a user's machine, enabling quick session resumption witho…
Z.ai released ZCode, a free desktop AI coding environment powered by its open-weights GLM-5.2 model, which benchmarks within one percentage point of Claude Sonnet 5 on coding tasks while undercutting …
A developer built an open-source tool that matches a person's resume to a Pokémon character using embeddings and a reranker. The project, hosted on Hugging Face Spaces, uses the BAAI/bge-m3 model to e…
Researchers built an external cognitive architecture called a "research world model" that allows an AI agent to record and query its trial-and-error experiences across contexts, enabling it to autonom…
Tessl developer Alan Pope built lincubate, an open-source tool that sandboxes AI coding agents like Claude Code and Codex inside LXD containers. The tool isolates agent configurations and skills per p…
A developer released Mindwalk, an open-source visualization tool that replays coding-agent sessions on a 3D map of a codebase, allowing users to see where an AI agent searched, read, and edited code. …
Google launched Stitch, a free AI UI generator that produces production-quality HTML/CSS or Tailwind from text prompts in under 60 seconds, with a March 2026 update adding MCP server support for direc…
Developer Piyush Khemka built Sip, a Claude Code status line plugin that estimates the water consumed by each AI prompt in real time. The tool displays per-response and daily water usage alongside cos…
An engineer at Anthropic has developed a multi-model playbook for Claude Code, arguing that the biggest leverage comes from loop engineering—designing agentic loops that use cheaper, faster models for…
A proxy experiment on Claude Code found that pruning stale file reads and unused tool schemas mid-session cut input tokens by 77% and cost from $2.04 to $0.48 on a JavaScript fixture, without degradin…
A developer released Confessor, an open-source tool that reconstructs the history of AI coding agents like Claude Code from local session logs to reveal every file read, command run, and secret access…
Developer Dan published a critical review of Overplane, an AI coding harness that uses LLMs and SMT solvers to generate verified code from Markdown specs. Dan argues the tool is redundant for experien…
A developer built the AI Agent Profiler, a transparent proxy that records every byte of traffic between AI coding agents and language models. Profiling Claude Code sessions revealed that agents re-sen…