Full-text search across 61544 articles. Combine with topic and date filters; results sorted by relevance.
A proposed build workflow for coding agents advises developers to construct interactive product demos around a single visitor decision, with explicit labels distinguishing real application logic, sample data, and simulat…
A new approach to evaluating AI coding agents, called behavioral evaluations, is gaining traction as a more reliable alternative to end-to-end benchmarks, according to a technical article on harness engineering. The meth…
General Robotics, an Nvidia-backed robotics platform, reported that AI wrote 80% of its code and reduced robot training time by 99%, according to a Forbes article by John Koetsier. The company claims agentic AI can drama…
A GitHub pull request by croviatrust proposes changes to the Causari AI coding workspace to validate snapshots before restoring and preserve the pre-bisect workspace. The audit reports 95.7% AI survival (1487 of 1554 lin…
Developer built AgentPulse, a tmux plugin that displays the status of Claude Code and Codex agents in the terminal multiplexer, using lifecycle hooks to show working, waiting, idle, and interrupted states. The tool adds …
A Hacker News post by user Marvin_RunAI reports that HAE (High Assurance Environment) has logged where the seconds go in a coding-agent turn, breaking down time into TTFT (time to first token), decode, and tools. The pos…
Simon Willison released a new AI code comment detector built on public data, replacing a previous version trained on private data that could not be shared. The tool runs entirely in the user's browser, and its interface …
A developer demonstrates how pairing screenshots with brief written notes gives AI coding assistants the visual context needed to make precise interface edits. The approach helps assistants understand layout, spacing, an…
SpaceX AI's Grok 4.6 is reshaping performance-per-dollar benchmarks, while coding agent valuations surge past $40 billion and neo-cloud infrastructure providers secure multi-year backlogs, according to an investigation i…
A developer argues that AI code reviewers claiming to follow team standards often fail to ground their feedback in actual rule files, instead relying on generic style guides and PR descriptions. The post advises testing …
Accordo, an open-source framework for building custom CRMs with coding agents, demonstrates a quote-approval workflow where an agent can request a discount but requires human approval above a threshold. The example shows…
Carlo Capocasa published a 10-task benchmark comparing coding agents Claude, OpenCode, Pi, Zcode, Hermes, and 3code on SWE-bench verified tasks, with 3code solving 9 of 10 tasks using 5 million tokens and Pi solving 6 of…
A developer has created an open-source system that gives AI coding agents persistent memory by indexing past conversation transcripts locally. The setup uses a FastMCP server connected to an SQLite FTS5 database, allowin…
Meta's AI agent Muse has entered a closed alpha phase with invite codes, moving beyond its previously announced waitlist, and is currently limited to Saint Kitts and Nevis. The agent will manage email, book dinners, trac…
Fatih Kadir Akın founded prompts.chat, which has become the world's premier open-source prompt engine by treating prompt strings as executable code for neural runtimes, anchored in Git version control, CSV data streams, …
GitHub resolved a service degradation affecting Copilot code review on September 4, after investigating reports of impacted performance for some GitHub services. The company identified the root cause and applied a mitiga…
OpenAI has announced GPT-6 Astra, its latest model, which it calls its most intelligent and aligned so far, claiming a new frontier in computer use, coding, and science.
Engram, an open-source shared procedural runbook memory for AI coding agents, launched on Hacker News, offering peer-verified execution-anchored procedures via MCP and REST, with Nostr-based identity and attestations. Th…
Agent-manager, an open-source tool by developer yoanwai, offers a keyboard-driven workflow for managing multiple coding agents such as Claude, opencode, codex, grok, gemini, pi, and hermes, allowing users to spawn, answe…
Frederik Wouters released a Claude Code skill that blocks em dashes from Claude responses, injecting a rule before every turn and forcing a rewrite if any of seven forbidden dash characters appear. The skill is installed…