A developer outlined a control-loop framework for coding agents that models each engineering episode as a state machine spanning intake, observation, modeling, planning, action, evaluation, and exit. …
A developer has released mcp-lint, an open-source Python design linter for Model Context Protocol servers that scores a server's tools/list output out of 100 and flags issues such as missing descripti…
Dipesh Sapkota, a computer engineering student at IOE Thapathali in Nepal, built Git Visualizer, an interactive React-based web tool that translates abstract Git commands such as commit, branch, check…
A developer outlined a process for converting AI code-review findings into deterministic guards, requiring that a finding be validated, assigned an owner, reproduced, encoded as an invariant, and prov…
A developer's experiment on a durable TypeScript and SQLite reminder queue found that both the Astra solo and Astra + Luna implementations passed their original acceptance suites while still containin…
BackBond scanned 7,749 distinct public MCP tool manifests from 8,147 reachable unauthenticated endpoints, covering 130,322 tool definitions, using its Agent Scan static metadata checker at ruleset 0.6…
Andrej Karpathy published an "idea file" gist, llm-wiki.md, describing a pattern in which an AI agent writes and maintains a linked markdown knowledge base from user-supplied sources, compiling synthe…
A member of the PinkWallet team outlined a design for an approval layer that sits between AI agents and company money, deciding per payment request whether an agent may pay. The proposed logic runs re…
A paper from UC Berkeley, TTIC, and Google Research authored by Keegan Harris, Siddharth Prasad, Asher Trockman, Nika Haghtalab, and Michael I. Jordan shows that sorting an LLM inference queue strictl…
A developer is building PlanToday, an AI-powered event and wedding planning platform that converts natural-language requirements — such as a wedding in Delhi for 300 guests on a ₹15 lakh budget — into…
Android developer Artem Garazha reports that Android 17's mandatory resizability rules remove legacy manifest opt-outs like screenOrientation and resizeableActivity for apps targeting API level 37 on …
A developer built cliffhanger, an open-source Stop hook and skill for Claude Code that blocks the agent from ending a turn while checklist items remain open, allowing a real stop only for named blocke…
A developer built Cyber Autopsy, a Kaggle benchmark that tests whether AI models can reconstruct reported cyber incidents from partial evidence while citing support for each claim and flagging what re…
Shopify extended its WebMCP protocol to checkout on September 28, 2026, letting browser-based AI agents read, update, and submit orders on eligible merchant storefronts with no merchant configuration,…
A developer argues that AI features should be designed with explicit degradation contracts, mapping each failure mode to a defined product behavior rather than letting a model outage disable an entire…
A developer built "Repair Without Breaking," a Kaggle Benchmarking Challenge dataset of 60 JSON episodes across 12 graph motifs and five matched conditions (applicable change, no-op, irrelevant update…
Apex Grid Technologies implemented a two-stage preflight check that verifies GPU accessibility and responsiveness before allowing GPU-dependent cron jobs to run, preventing silent failures and wasted …
DIVD reported that its own Zammad instance was compromised through a chain of two zero-days, CVE-2026-102489 for session hijacking and remote code execution as the zammad user and CVE-2026-102490 for …
A developer built Forge, an open-source full-stack STEM copilot that streams step-by-step derivations over Server-Sent Events and runs a second, independently configurable model as a cross-check, surf…
A developer built a three-stage deploy gate that lets a fully autonomous Claude Code-based coding agent merge and ship its own pull requests safely, after an early incident where a currency-blind cach…