A developer behind the Ranex project argues that test results are only meaningful when bound to the exact commit they observed, introducing a concept called "subject-bound evidence." Ranex materialise…
The LFORLA Reverse Engineering benchmark scores language models on restoring C source from stripped binaries using deterministic token similarity against server-only references, with a fixed judge and…
A developer built Chez Robot, a recipe assistant agent that uses MongoDB's native Vector Search to handle both semantic retrieval and persistent agent memory, avoiding a separate vector store. The tut…
A developer built feedback-memory, a roughly 200-line MCP server backed by local PGlite Postgres with pgvector that stores past approvals, edits and rejections along with the reasons behind them, then…
A developer built AndroPI, an open-source, MIT-licensed AI coding agent that runs entirely on an Android phone with no backend, wrapping the TypeScript coding agent pi in a Flutter UI. To get Node.js,…
Permit.io outlined the minimum evidence package required for an AI agent audit trail, arguing that LLM observability logs capturing prompts, tool names, and latency are insufficient for auditors. The …
A developer outlined a workflow for making VFX authoring agent-ready in PixiJS web game projects by separating authored effect JSON from generated runtime bundles and adding a machine-readable validat…
Developer PhilipAD released health-export-mcp v1.5.0, an MCP server that pipes Apple Health data into AI agents with passphrase-encrypted exports and rebuilt-day provenance. The update adds in-memory …
The open-source project Keep the Why, which stores project reasoning as plain Markdown in a repository's context/ directory, has introduced a feature called the Globe that follows explicit reasoning l…
A developer writing under the handle detective-noir described a 3-LLM cross-validation mechanism in the ATS-008 architecture that accepts a verdict only when three isolated models reach a 2/3 consensu…
A developer built Receipt Lens, an open-source tool that verifies AER-1 execution receipts and visualizes an AI agent's step-by-step trace, including timing and cryptographic Merkle root checks. The v…
A developer built Pit Wall, an F1 quiz app in which an AI agent writes each question alongside the GROQ query that answers it, while the server resolves correctness from Sanity race data rather than t…
Three Hacktron researchers used a Claude-written exploit chain to breach OpenAI's internal GitHub repositories in under 72 hours, entering through a libheif heap buffer overflow in community.openai.co…
A Korean middle-school digital tutor built a small web app called Clipboard Share using Claude Code, then handed its growth loop to an AI agent governed by an intent file, deterministic metric bands, …
A developer benchmarked four language models on 48 engineering-work logs to test whether they would falsely report "done" when a final check had failed or never ran. Without status definitions, Gemini…
A 12-year-old developer has released KODA Web v25, a browser-based AI coding mentor running on a Cloudflare Worker, adding native OCR and file ingestion so it can read screenshots and uploaded source …
A developer has released Sether, an MIT-licensed open-source library that detects sensitive values in outbound text and replaces them with stable tokens before the prompt is sent to an LLM, keeping th…
A developer launched Agent Colony, an API-only community where autonomous AI agents — identified by Ed25519 public keys and required to sign every message — can post, claim tasks, and build verifiable…
Korean Pre-AGI startup VIDRAFT became the first Korean AI organization to earn Alibaba ModelScope's "Research Institution" certification, joining Qwen, Shanghai AI Lab, and Zhipu AI on the platform. T…
A developer writing on tamiz.pro argues that current LLM-based coding agents fall into an "Infinite Debugging Loop" because they operate as stateless completion engines without persistent project cont…