---
title: "Anthropic Engineer Discusses AI Agent Verification"
url: https://stacklist.com/card/b1886843-3c2e-4ad4-ae39-433bc2466dc9
source_url: "https://x.com/0x_rody/status/2063318596202242171"
stack: https://stacklist.com/stack/539051c9-e760-4bd5-a86b-ac120e4ae368
summary: "Anthropic engineer James Brady discusses how every AI agent in production lies, sharing a verification stack and 4-layer setup to make hallucination expensive. The talk and accompanying article detail patterns adopted by the Claude Code team to catch fabricated functions, fake imports, and false test results."
tags: "ai-agents, truthfulness, claude-code, hallucination, verification, anthropic, llm-reliability"
key_entities: "James Brady (person), rody (person), Anthropic (organization), Claude Code (technology), AI agent truthfulness (concept), verification stack (concept), hallucination mitigation (concept)"
classification: "snippet"
content_hash: "sha256:70643e05334d28576d7ab470bcb8c17c6ed2fdc6124c8500102d4792739c442f"
acp_version: "0.2"
token_counts_approximate: 727
visibility: public
agent_accessible: true
status: "final"
---

# Anthropic Engineer Discusses AI Agent Verification

rody @0x_rody Anthropic engineer James Brady: "Every agent in production lies. We measured it. The good ones lie less, the great ones catch the lie before the user does." In 29 minutes, he walks through the verification stack he built and the patterns the Claude Code team adopted to keep Watch the full talk, then save the config below👇 00:00 rody @0x_rody Jun 6 Article How to Make Claude Code Stop Making Stuff Up When It Doesn't Know (Exact Setup Inside) Claude Code lies to your face every day. Made-up functions, fake imports, "tests pass" when nothing ran. The fix is a 4-layer setup that makes lying expensive. Claude tells you the bug is fixed. Two... 5:54 PM · Jun 6, 2026 340.4K Views :host{display:inline-block;direction:ltr;white-space:nowrap;line-height:var(--number-flow-char-height, 1em) !important}span{display:inline-block}:host([data-will-change]) span{will-change:transform}.number,.digit{padding:round(nearest, calc(var(--number-flow-mask-height, 0.25em) / 2), 1px) 0}.symbol{white-space:pre} 4 3 number-flow-react > span{font-kerning:none;display:inline-block;line-height:var(--number-flow-char-height, 1em) !important;padding:calc(round(nearest, calc(var(--number-flow-mask-height, 0.25em) / 2), 1px) * 2) 0} 43 :host{display:inline-block;direction:ltr;white-space:nowrap;line-height:var(--number-flow-char-height, 1em) !important}span{display:inline-block}:host([data-will-change]) span{will-change:transform}.number,.digit{padding:round(nearest, calc(var(--number-flow-mask-height, 0.25em) / 2), 1px) 0}.symbol{white-space:pre} 1 6 0 number-flow-react > span{font-kerning:none;display:inline-block;line-height:var(--number-flow-char-height, 1em) !important;padding:calc(round(nearest, calc(var(--number-flow-mask-height, 0.25em) / 2), 1px) * 2) 0} 160 :host{display:inline-block;direction:ltr;white-space:nowrap;line-height:var(--number-flow-char-height, 1em) !important}span{display:inline-block}:host([data-will-change]) span{will-change:transform}.number,.digit{padding:round(nearest, calc(var(--number-flow-mask-height, 0.25em) / 2), 1px) 0}.symbol{white-space:pre} 1 . 7 K number-flow-react > span{font-kerning:none;display:inline-block;line-height:var(--number-flow-char-height, 1em) !important;padding:calc(round(nearest, calc(var(--number-flow-mask-height, 0.25em) / 2), 1px) * 2) 0} 1.7K :host{display:inline-block;direction:ltr;white-space:nowrap;line-height:var(--number-flow-char-height, 1em) !important}span{display:inline-block}:host([data-will-change]) span{will-change:transform}.number,.digit{padding:round(nearest, calc(var(--number-flow-mask-height, 0.25em) / 2), 1px) 0}.symbol{white-space:pre} 4 . 8 K number-flow-react > span{font-kerning:none;display:inline-block;line-height:var(--number-flow-char-height, 1em) !important;padding:calc(round(nearest, calc(var(--number-flow-mask-height, 0.25em) / 2), 1px) * 2) 0} 4.8K Read 43 replies