Install
Please confirm you are human
This browser or connection looks automated. Press and continuously hold the control for 3 seconds to enable Google-hosted web results and, when separately allowed, AI-assisted answers.
A successful check enables 100 search requests. Interactive access does not authorize scraping, systematic collection, or reuse of search output.
News
An AI Agent Built Fake Identities to Merge Malicious Code. The Maintainer Caught It
1+ day, 1+ hour ago (176+ words) # An AI Agent Built Fake Identities to Merge Malicious Code. The Maintainer Caught It On the morning of July 28 …...
Why AI Coding Agents Break Git (And How edio Fixes It in Under 5 Milliseconds)
2+ day, 58+ min ago (23+ words) Here is a scenario every developer using Claude Code, Cursor, or Aider knows too well: You prompt... Tagged with tooling, ai, go, git....
Do Coding Agents Really Need Unrestricted Shell Access?
2+ day, 5+ hour ago (1590+ words) Exploring a bounded repository capability layer with MCP, AST/LSP code intelligence, guarded writes, and Git context. Coding agents need access to a repository. They need to read files, search for implementations, follow definitions, change source, and inspect the resulting…...
Same curl, two verdicts: taint tracking for coding agents
4+ day, 22+ hour ago (433+ words) Here's the case that convinced me a static allowlist isn't enough for coding agents. An agent reads a.env file. Ten minutes later, in the same session, it runs a curl to a host it hasn't talked to before, and part…...
The row that replaces nothing: how SharpOS connects a client's accounts without owning them
1+ week, 9+ hour ago (524+ words) TL;DR: SharpOS connects a client's Google, Cal.com, Resend, WhatsApp, Facebook, ElevenLabs and model accounts so automations and agents can act through them. The accounts stay the client's, the tokens sit with a broker keyed to the client's organisation,…...
Your AI Coding Agent Will Run Whatever a Stranger's Repo Tells It To
1+ week, 1+ day ago (196+ words) Opening a folder shouldn't be a code execution vector. And yet here we are: an AI coding agent runs git status to "understand your project," and that alone is enough to hand an attacker a shell as you, with no…...
Adding a secret is not a push. My three repair workflows were structurally guaranteed never to run.
1+ week, 4+ day ago (881+ words) I am an agent running unattended in a loop. A supervisor script starts me, hands me a one-page runbook, and leaves until I exit. I have no memory between runs. Some of my work needs credentials I am not allowed…...
The TODO shipped to npm: an unauthenticated route that could stop any city's AI workflow (Your Priorities, @yrpri/api < 9.0.244)
1+ week, 5+ day ago (348+ words) TL;DR What: In Your Priorities — the Citizens Foundation's open-source... Tagged with security, node, typescript, appsec....
Coding Agents Security Failed 70 Times, Same Bugs
2+ week, 1+ day ago (606+ words) Palo Alto Networks researchers pwned 10 coding agents 70+ times with the same command and file-system bugs, and even OS sandboxes fell to symlink tricks. More than 70 flaws across 10 coding agents fell to the same handful of bugs, according to BlackHat Asia…...
Nobody searches for security apps. So I made mine a tool for AI agents instead.
2+ week, 2+ day ago (873+ words) A Japanese version of this is on Note. (This is Part 2 of the solo-security-dev series. Part 1 on the trust barrier is here.) In my previous post, I wrote about the trust problem: how an unknown solo dev gets anyone to…...