Full-text search across 68438 articles. Combine with topic and date filters; results sorted by relevance.
Earendil shipped Pi 1.1 on October 7, adding MCP support to the MIT-licensed coding agent nearly a year after Pi creator Mario Zechner said on November 30, 2025 that Pi would never support MCP, citing the roughly 18,000 …
Developer Enes Alatas built a custom merge queue in GitHub Actions for the personal repository hosting Spool, a tool for creating and managing persistent Claude Code loops, after AI agents generating pull requests made G…
Paul Ford, writer and programmer behind Businessweek's "What Is Code?" issue, launched Tiny Little Office on October 14 through his software firm Aboard, a free ten-app office suite built with Claude in roughly ten days …
Developer aldohushi1-stack released Shipback 0.1, an open-source Node 18+ command-line tool and MCP server that deploys a folder to any SFTP host — cPanel, Plesk, DirectAdmin or a plain VPS — and backs up every file it o…
Past.dev launched a long-term memory service for AI agents that, in a six-task Claude Code+ comparison published by the company, produced correct answers on all six tasks versus one correct, four wrong and one not-done r…
Perplexity open-sourced SPACE (Sandboxed Platform for Agentic Code Execution), the Firecracker microVM sandbox behind every Computer session since July, on October 15 under Apache 2.0. The Rust-built system, which runs m…
Giang Pham, maker of Pianory, built a score importer that validates uploads by content rather than filename, detecting ZIP-based .mxl archives by byte signature and resolving the actual score through META-INF/container.x…
A developer built an open-source candidate screening call agent on Telnyx that places outbound AI-conducted phone interviews and persists each candidate's profile, answer ledger, scores and retry history in a durable per…
Security researcher Paulos Yibelo of Pwn.ai discovered PwnEscape, tracked as CVE-2026-99999, a Linux kernel networking flaw that let an attacker inside a virtual machine alter an IPv4 header after the kernel had processe…
A developer built CleanSync.ai, an open-source, fully local security airlock that intercepts files downloaded from browsers, messaging apps, and cloud storage into an isolated staging folder before they reach primary sto…
Software engineer kqr tested James Shore's claim that a month of new code costs ten days of maintenance in its first year and five days every year after, fitting a geometric decay model to eight open-source project histo…
A developer submitted HospitalFlow AI, an open-source web dashboard that flags hospital department bottlenecks as Normal, Warning or Critical by comparing workload against staff capacity, to the Hacktoberfest Open-Source…
VLLM supports Reflection's Beam, a 501B-parameter mixture-of-experts model with 23B active parameters per token, from day 0, with the model served through the Docker image vllm/vllm-openai:reflection-beam. Coding agents …
A developer who contributed to GitLab Core's Model Context Protocol implementation benchmarked MCP against GitLab's traditional REST API v4 for autonomous AI coding agents, finding MCP resources cut raw payload size by 9…
Reflection AI's Beam, a low-latency US open-weight reasoning model for coding and agentic tasks, is now available on Vercel's AI Gateway and free to use for a limited launch period. Beam activates only 23B of its 501B pa…
The TypeScript compiler team merged a change that embeds AST node headers directly at offset zero within concrete node structures, shrinking header size from 48 bytes to 32 bytes and eliminating a separate self-referenti…
A developer built Aria, a conversational booking assistant for Prime Coworking that owns no availability or booking logic of its own, instead wrapping the 15 domain services the product's existing booking widget already …
Llm-gauze, an HTTP gateway published as the container image ghcr.io/ggeorgovassilis/llm-gauze:latest, sits in front of a local OpenAI-compatible LLM and remediates transient errors, hung or looping models, empty response…
VLLM shipped day-0 support for Muse Spark Open, an open-weight sparse mixture-of-experts model built with Meta Superintelligence Labs for long-horizon agentic coding workflows, carrying 816B total parameters, 42B active …
Vercel added Cat Jam, a reasoning model for coding and agentic tasks with low-latency inference, to its AI Gateway as a stealth model that is free to use during the stealth period. Cat Jam supports function calling, stru…