Is this you? As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.
Claim your profile
Articles by Pushpendra Agrawal
AI Agent Circuit Breakers
AI Agent Circuit Breakers You deploy your shiny new autonomous research agent on Friday afternoon. It works perfectly on your curated test data. It navigates to a URL, extracts the core thesis, cross-references it against a vector database, and generates a clean JSON summary. You close your laptop and enjoy your weekend. You check your OpenAI or Anthropic billing dashboard on Monday morning. Your weekend staging environment cost is sitting at $542.18. What happened?
When a Failed Request Must Stay Failed: Reservation Replay
This is a submission for the Kaggle Benchmarking Challenge. What I Benchmarked A room is occupied, so a booking request fails. The room becomes free. Should an identical retry now succeed? Under this benchmark's declared contract, no. The original request ID must replay the original rejection. A genuinely new attempt needs a new ID. This is the rule being tested, not a claim about every reservation API.
I Built a Local-AI Character Forge and a Browser Game Engine on a $100/y VPS (No GPU, No Excuses)
There's a special kind of stubbornness that leads someone to say "I will generate photorealistic 3D humans, dress them with AI-designed clothing, give them personalities via an LLM, and drop them into a playable FPS/RPG hybrid — on a VPS with 6 vCPUs, 11 GB of RAM, and zero GPU." That stubbornness is basically the whole architecture document for this project.
Why an AI Agent Can Execute the Same Action Twice
AI agents are becoming execution systems. They no longer just answer questions. They send messages, create tickets, issue refunds, make bookings, update customer records, trigger deployments, provision resources, and call tools that change external state. That creates a failure mode distributed-systems engineers already know well — but agent loops make it unusually easy to trigger: a tool call can succeed and still look like a failure to the agent.
Not Another LLM: I Tried Laya, a 421M-Parameter AI Decision Engine
I spend a lot of time trying different AI models, especially models that can run locally. Usually, the process is pretty familiar: download the model, give it a prompt, and wait for it to generate something. Laya caught my attention because it doesn't work that way. It isn't built to write an answer, generate code, or have a conversation. Instead, you give it some text or a piece of information and ask it specific questions. It then returns a decision and a probability for that decision.
1Presence: What We Built and How It Works
Originally published on the 1Presence blog on 27 August 2026, the day we launched. Jon here, founder of 1Presence. We launched today. This is the long version of the launch note: where the product came from, what it does, and how it is built. Where this came from I worked with a brilliant team at MIT on the AI chat interfaces for its online learning platform. I built the agentic backend for Senis AI, an intelligence platform for membership associations, end to end.
I built a public, cryptographically signed test registry for AI agents (student project)
Hi everyone! ð Iâm a final-year BTech CSE student. The Problem I Noticed As AI agents start doing real tasks (like sending emails or accessing files), we currently just have to trust the developer's word that the agent is safe from being tricked or hijacked. There is no simple, public way to verify an agent's actual safety record. My Attempt at a Solution: A2A-G I built a small, open-source project called A2A-G to explore if we could make agent safety public and testable.
Forms that lose leads silently
A contact form can look perfect and still throw away every lead that comes through it. The visitor fills the fields, clicks submit, sees a thank you message, and leaves satisfied. Nobody on either side knows the message never arrived. This is a different failure than the one everyone talks about, the confusing layout or the field that asks for too much. Here the form works, in the sense that it renders, accepts input, and reports success. It just doesn't deliver.
The agent host didn't have the lifecycle boundaries I assumed
As I started handing agents longer pieces of work, I stopped being present at every lifecycle boundary. In shorter copilot-style sessions I had been part of the control loop without really thinking about it. Longer agent runs broke that assumption: I was no longer present to know when work finished, what mattered, or what context should survive into the next task. One place this surfaced was memory.
Plugin4Shell: When Your AI Coding Agent Auto-Updates Straight Into RCE
Zero-click RCE. Four major AI coding agents. No user interaction required. Let's talk about Plugin4Shell. The Incident In September 2026, researchers disclosed a vulnerability class dubbed "Plugin4Shell" affecting Claude Code, Codex, Gemini CLI, and Copilot. The flaw lives in how these agents verify SHA-pinned plugin commits pulled from marketplaces.
5 failure modes of autonomous coding agents and how to catch them
Autonomous coding agents look great in demos. They read a ticket, write code, run tests, open a PR. Then you put one in a real repo with real credentials and real users, and it starts doing things nobody asked for. Most of the incidents I've seen (and caused) trace back to five recurring failure modes. None of them are exotic. All of them are catchable if you build the right checks before you ship, not after something breaks. 1.
FraudLens - TigerGraph
Building an Autonomous GraphRAG Fraud Investigation & Next-Best-Action Agent with TigerGraph A deep-dive into graph-native financial crime investigation, evidence-sufficiency gating, and self-updating institutional memory. Author: Team Hacker House Goa 2026 Stack: TigerGraph Savanna · GSQL & Native Algorithms · Model Context Protocol (MCP) · LangGraph · Claude / Anthropic · FastAPI · React & Tailwind 1.
Your AI agent has more permissions than your users
Here is a conversation that happens in a lot of companies right now: Dana: @assistant please close PAY-123 and delete the old release branch Assistant: Done ✅ The problem: Dana is not allowed to delete branches in that repository. The assistant is. Most AI agents and chat bots act in other systems (Jira, GitHub, Slack, Salesforce, AWS) through one service account. That account needs enough access to help everyone, so it ends up with more access than any single person who talks to it.
Why RBAC Alone Isn't Enough for Enterprise Data Agents
A user can be blocked from a sensitive column and still receive sensitive information derived from data they are allowed to access. That changes the authorization problem for enterprise data agents. Traditional access control asks: Can this user read this database object? An AI analytics system also needs to ask: Is this user allowed to receive what the system can infer from those objects? Consider a simple example.
What to verify after an AI agent says tests pass
Green CI from an agent is not a merge signal. It is a claim: "I ran something and it exited zero." Your job is to check whether that claim covers the bug, the intent, and the failure modes you care about. I treat "tests pass" as the start of a short verification sequence — not the end of review. 1. Inspect the diff, not the summary Open the file list first. Ignore the agent's narrative until you can answer: Which files actually changed? Do they match the human intent in one sentence?
A Practical Checklist for an Agent-Ready API
An MCP server exposes tools. It does not repair an API that leaves side effects, retries, data limits, and recovery ambiguous. Use this checklist before exposing an endpoint to an agent. Classify the effect Every tool should identify one effect: read, draft, state_change, or irreversible_action. The calling layer, not the model, should enforce approval for consequential effects.
‘It Answers in Poetry Now’: A Week of Users Saying Their AI Got Wordier and Worse
For months the loudest complaints in consumer AI have been about money and access: the meter, the rate limit, the price rise, the paid feature that quietly vanished. This week the grievance moved somewhere less obvious and more interesting — the writing. Across the big subreddits, heavy users of Claude, Gemini and ChatGPT converged on the same complaint, in almost the same words: the models have started sounding cleverer and answering worse.
Open or closed model? There is a third option
Open or closed? The question shows up in every architecture review, and it usually gets settled with a benchmark chart and a price-per-token comparison. But it often gives the wrong answer, because the two options on the table are not the two you have. There is a third sitting between them, and for most workloads it wins. Why the binary is badly posed The usual reasoning runs like this. An H100 rents for a few dollars an hour, while a frontier API charges a markup per token on that same hardware.
Show More
loading
Actions
Is this you?
As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.Get in touch with Pushpendra
Contact Pushpendra, search articles and posts on X, monitor coverage, and track replies from one place.
Learn more about Muck Rack