SShortSingh.
TopTechnologyAI & MLStartupsProgrammingScienceBusinessFinanceCrypto & Web3World
Back to feed
Programming

Claude Parallel Tool Use: My Agent Loop Crashed on 388 of 1,200 Runs

0
·1 views

My codebase Q&A agent ran 1,200 times in its first week. 388 of those runs died with the same HTTP 400. The other 812 worked fine, which is the worst possible ratio: too low to look like a broken deploy, too high to ignore. The cause was one line I'd copied from my own earlier prototype: next(b for b in resp.content if b.type == "tool_use"). It grabs the first tool call.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Log in to join the discussion and vote.

Log in

Related stories

0
ProgrammingDEV Community ·

How to See Any Company's LinkedIn Ads with Python (Ad Text, Impressions and Targeting)

If you sell to businesses, your competitors are almost certainly advertising on LinkedIn: webinars, whitepapers, demo offers, "thought leader" posts from their executives. LinkedIn lets you see all of it. The LinkedIn Ad Library lists every ad that ran on LinkedIn in the past year, and for ads shown in the EU it even adds run dates, impressions, the country split and the targeting the advertiser chose (the EU's Digital Services Act requires it). The catch: it's a search page for humans. No export, no API, 24 ads at a time, and a name search that's fuzzy.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Our Permissions-Policy denies geolocation, and the pricing page still knows what country you are in

curl -sI https://pub-trivia.app/ permissions-policy: camera=(), microphone=(), geolocation=() referrer-policy: origin-when-cross-origin strict-transport-security: max-age=63072000; includeSubDomains; preload x-content-type-options: nosniff x-dns-prefetch-control: on x-frame-options: SAMEORIGIN No x-powered-by. Six headers, declared once in next.config.mjs for /:path*, and each one is a decision rather than a default. The two most interesting are the ones that look like they contradict the product. The pricing page shows you a price in your own currency. It knows roughly where you are.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Getting to Know Yourself: Understanding self in Ruby

self is one of those Ruby keywords that looks easy until you meet it in real code. You see it in a setter: self.name = normalized_name Then in a class method: def self.from_cents(...) Then somewhere inside a DSL or instance_eval, and suddenly the question isn't really: What does self mean? The better question is: Who is self right here? That small change makes Ruby's object model much easier to follow. Consider a model-like object: class Customer attr_accessor :name, :email def initialize(name:, email:) @name = name @email = email end def normalize self.name = name.strip.downcase self.email =

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

robots.txt matching is a literal prefix, and one trailing slash would have left our dashboard crawlable

Our robots.txt is generated from one array: export const CRAWLER_DISALLOW = [ '/dashboard', '/play', '/api', '/auth', '/subscribe', '/webhook', '/too-many-requests', ] as const Not one of those has a trailing slash, and that is the first thing I would check in anybody else's robots.txt. Disallow is a literal prefix match against the path. Disallow: /dashboard/ blocks /dashboard/settings and /dashboard/billing, and leaves the bare /dashboard perfectly crawlable, because /dashboard does not begin with /dashboard/. In our case that is not a hypothetical: /dashboard is linked from the footer of th

0 comments Read more at DEV Community