SShortSingh.
TopTechnologyAI & MLStartupsProgrammingScienceBusinessFinanceCrypto & Web3World

Programming

LatestTopDiscussed
0
ProgrammingDEV Community ·

Save your main context to raise quality and cut cost — building Context Drop

1. Hook + thesis At some point, Claude Code started to "break" on me. To be precise, what broke wasn't Claude Code itself. What broke was the way I was using it. Firing off dozens of agents at once, pasting giant logs and screenshots straight into the conversation, keeping long sessions alive by stitching them together with /compact — the heavy, fan-out-centric style that Claude Code's ultracode opt-in naturally produces.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Griha: the chore ledger where the fairness argument is auditable, and TabPFN runs in your tab

Entry for the Hacktoberfest Weekend Challenge — Build for a Friend (#hf26challenge) Best Use of TabPFN (Prior Labs) category. Someone in a shared home who has quietly become the household's project manager. Not the manager by title — nobody appointed them. They are the one who remembers the geyser vent needs cleaning, who notices the milk is out, who ends up asking "did you take the recycling out?" in a tone that makes everyone hate them. Then, when the work actually gets done, nobody can say whether it was fair.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Let the model decide — then read the audit

Most "governed agent" demos are governed because the demo author pulled every string. The agent calls the tool the script says, in the order the script says, and the audit comes out clean because nothing could have gone another way. That proves the plumbing. It does not prove governance, because governance is what happens when the thing you are governing has a will of its own. So we took the strings off.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Claude Code Mods Just Turned Agents Into a Platform

Claude Code Mods Just Turned Agents Into a Platform On October 1, 2026, Anthropic shipped Claude Code v2.1.287 with a feature called Mods — JavaScript and TypeScript functions that hook into the agent's execution pipeline, intercept tool calls, render custom UI, and reshape how the agent behaves from the inside out. Boris Cherny, the head of Claude Code, announced it on X to 833,000 views and 5,200 likes. Within hours, someone had Tetris running inside their terminal. Read the full version with charts and embedded sources on AgentConn → View original post on X → The same week, @AGTPinsights pu

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Flash Loan Attack Vector Analysis: Gauntlet

Flash Loan Attack Vector Analysis: Gauntlet Target Protocol: Gauntlet (TVL: $1628.6M) Flash‑Loan Attack Vector Analysis – Gauntlet Protocol: Gauntlet (TVL ≈ $1.63 B across Ethereum & L2s) Date: 4 Oct 2026 Prepared by: Senior DeFi Security Researcher – Smart‑Contract Audit Team Gauntlet is a strategic‑simulation and risk‑management platform that supplies on‑chain data, capital‑allocation recommendations, and automated execution for large‑scale DeFi investors. Its core value proposition is the ability to run “what‑if” simulations using real‑time market data and then execute multi‑step, multi‑pro

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

I built my husband a vim trainer with a Gemma coach that runs in the browser

This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend A lot of my friends build their own tools. They hit a problem and just make the thing they want. I wanted to build something for someone else this time, and the answer was sitting right next to me. My husband has wanted to learn vim for a long time. He's tried a bunch of times, and it always ends the same way: he gives up.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

A customer maintenance form in Uniface 10, part 3 - a service, 30 automated tests, and the bug that came with them

At the end of [part 2] the customer form worked and contained everything: e-mail rules, SQL strings, quote escaping, a number range, a duplicate query. All of it inside a form component, reachable only by a human clicking buttons. This part does three things: Move the rules into a service (CUSTOMER_SVC). Write a test service (CUSTOMER_TST_SVC) that exercises them automatically and leaves the database untouched. Fix the bug that the refactoring exposed - 0129 - Error on field CUSTOMER_ID; subfield(s) are required - which took two attempts, because the first fix only moved it.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

New Benchmark Catches AI Agents Lying About Finished Work

A joint blog post from Microsoft and Hugging Face describes a new agent evaluation framework, ThinkingBox, built around a simple idea: grade an AI agent by what it actually wrote to a database, not by what it said in its final reply. The motivating example is a support agent handling a late delivery. It makes nine tool calls, reads the refund policy correctly, and closes the ticket as resolved. Two things are still wrong: the courier exception is still open, so the ticket should have been left on hold, and the customer never got an answer to her actual question. A grader checking only tool cal

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

VivaMate: A Local AI Interview Partner I Built for a Friend

This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend [Friend's name] has campus placements coming up and kept saying the same thing: "I know the answers, I just freeze when someone asks." Reading notes doesn't fix that. Practicing out loud with someone who pushes back does. So I built VivaMate, an AI interview partner that: Runs mock, HR, DSA and project deep-dive interviews, one question at a time Reads [Friend's name]'s resume PDF and GitHub repos, so the questions are about their projects Scores every answer for correctness and clarity, then tracks weak topics, s

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

My Own Monitoring Environment Wasn't "Outside" Either

Today's smallest, most concrete lesson came from trying to verify a suggestion before replying to it, and failing in a way that turned out to prove the point. The suggestion On a thread about a canonical-URL bug we shipped a fix for, a commenter pointed out something easy to miss: a check that validates a site's canonical tags needs to run from outside the site's own network, because CDNs can serve different responses to requests they recognize as infrastructure versus requests they treat as an ordinary visitor. A check running from inside your own deploy pipeline might pass cleanly while the

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

A customer maintenance form in Uniface 10, part 2 - the four things that turn CRUD into a business application

In [part 1] the customer form does the full CRUD cycle: search, select, edit, save, delete. It compiles, it runs, it looks fine in a demo. Then you hand it to someone who actually works with it, and within ten minutes you get four bug reports: "I created a customer and it overwrote another one." "I typed a phone number, clicked the next row, and my change was gone. No warning." "Searching for meier finds nothing, but Meier does. And I can't search by e-mail." "We now have the same customer three times." None of these are Uniface problems.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

A signed cookie is not object authorization

Signed cookies are great for integrity: the browser cannot silently rewrite a payload you sealed with a server secret. That still does not decide which objects the user may touch. Teams often stuff a user id, a role, or even a tenant id into a signed cookie and then treat “signature valid” as “request authorized.” The cookie only proves the blob was minted by you. It does not prove the caller may read invoice #4821, mutate another user’s project, or act across tenants. What to do instead: Authenticate the session (cookie, token, whatever) to learn who is calling.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Your 'vectorized' SVG may still be a bitmap — 4 checks, with measurements

When you "vectorize" an image, the result is supposed to be paths — geometry you can scale, recolor, and cut. A surprising number of tools, including Inkscape's own Trace Bitmap, instead wrap the original bitmap in an tag, so the file is a picture inside a vector wrapper. It looks fine at 100% and falls apart everywhere else. You can tell which one you got in about a minute. Here are four checks, then the numbers from a benchmark I ran.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

The Exemption Cliff: Why a 10% Policy Loophole Eats Most of the Backstop

Part 23 of the Fat Tail Notes series — the last garment's loophole is the policy's. Part 22 priced the retail loophole: under full commitment, an exemption rate of just 10% destroyed 85% of the garment, because as long as any override can fire, the worst 1% stays deep. The conclusion for the individual was clean: make the hole narrow. But the individual's hole only eats her own underwear. The backstop itself — the thing every portfolio leans on in a crash — has its own override, and it is the most dangerous one in the series.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

My Friend Kept Getting OAs but No Interviews. So I Built PlaceTrace.

This is a submission for the "Hacktoberfest Weekend Challenge: Build for a Friend" (https://dev.to/challenges/hacktoberfest-weekend-2026-10-01) What I Built A few weeks ago, my friend Kunal was going through placement season. He had been applying to different companies, taking online assessments, and waiting for interview calls. But after doing this again and again, he kept facing the same problem: OA → wait → no interview call. The frustrating part wasn't just getting rejected. It was not knowing where his placement journey was actually breaking.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

My Friend kunal Kept Getting OAs but No Interviews. So I Built PlaceTrace.

This is a submission for the "Hacktoberfest Weekend Challenge: Build for a Friend" (https://dev.to/challenges/hacktoberfest-weekend-2026-10-01) What I Built A few weeks ago, my friend Kunal was going through placement season. He had been applying to different companies, taking online assessments, and waiting for interview calls. But after doing this again and again, he kept facing the same problem: OA → wait → no interview call. The frustrating part wasn't just getting rejected. It was not knowing where his placement journey was actually breaking.

0 comments Read more at DEV Community
← NewerPage 664 of 3152Older →