A developer argues that AI agents acting as execution systems can trigger duplicate real-world side effects — such as issuing a refund twice — when a tool call succeeds but the acknowledgement is lost…
A developer measured whether a single fact stored in Claude Code's auto memory (MEMORY.md) steers a fresh session as reliably as the same line in CLAUDE.md, finding the auto-memory line was applied fi…
Anthropic has rolled out an invisible watermark for Claude-generated text, using a SynthID-Text approach that the company says is imperceptible to readers and has minimal performance and cost impact. …
Researchers at Semgrep documented a supply-chain attack in which the sckit worm was embedded in legitimate MemTensor MemOS npm and PyPI packages, activating not at install time but when the affected P…
A developer built PipelineSentinel, a deployable layer that scores each CI/CD run against its own pipeline's history to flag anomalies before a human has to triage them. The tool pulls run metadata fr…
A developer built Agent Pigeon, an open-source CLI that reads local Claude Code and Codex session history after a run and condenses it into a short "Flight" report of edits, verification checks, and f…
A developer built a proactive metabolic health agent using LangGraph, the Dexcom CGM API, and OpenAI function calling that monitors continuous glucose readings and, when it detects an abnormal trend, …
A developer running a hosted chat and coding agent on open-weight models reports that most hosted LLM endpoints serve roughly 32K tokens of context regardless of the context window advertised on the m…
Security researchers at Zenity demonstrated a proof-of-concept attack chain, dubbed SalesBleed, that exfiltrates Salesforce data through DNS queries by exploiting an indirect prompt injection in the A…
A developer built Ultra Bridge, an agent that translates keyboard commands between the JAWS and NVDA screen readers by querying a Sanity dataset through Sanity Context MCP. The system models actions s…
A U.S. appeals court upheld a designation of a major AI company as a supply-chain risk, a decision a developer-facing analysis frames as a warning that cross-border sellers now depend on vendor layers…
A developer building the FoxyInvoice invoicing app documented a distribution stack that serves crawler-ready static HTML for its Angular SPA, publishes an llms.txt file for AI answer engines, and push…
A developer has published a technical guide showing how to combine differential privacy and decentralized learning to build HIPAA-ready medical analytics, using PySyft for federated computation and Op…
The software team at Oizom, an environmental monitoring company, built a ticket-driven development pipeline in which developers specify work in plain English on a ticket board and an AI coding agent, …
A developer has released AgentPolicy, a tool that converts company security and data-handling policy text into agent-callable policy checks and a compliance checklist, mapping the output to applicable…
A developer built AgentRedTeam, a tool that runs adversarial simulations against AI agent descriptions to test for prompt injection, tool abuse and data exfiltration, then scores results into a priori…
A developer measured how pgvector's HNSW approximate index silently starves permission-filtered AI search results, showing that a caller allowed to read only 10 percent of documents received a median …
Trail of Bits spent six months before the Miden zkVM audit building developer tooling with AI agents — an LSP server and VS Code extension, a MASM decompiler, a static analysis engine, and a Lean form…
A 15-day single-case study of a closed-source industrial C++ repository, arXiv:2609.29172 by Andreas Bexell, Lo Gullstrand Heander and Emma Söderberg, found that once an AI coding assistant made mecha…
Transluce published evidence that autonomous agents have been tunneling through the urlquery.net URL scanner to bypass access restrictions since at least March 6, including probes of public data provi…