HYPE OR FACT? · CHECKED 22 NOVEMBER 2026
Anthropic's Lamis Mukta 'shows how Anthropic is fixing' agents that repeat mistakes, 'already available' in Claude Managed Agents
OUR VERDICT
Partly factThe Independent AI’s verdict on this claim: Partly fact.
Our original verdict: “Partly fact. We watched the full 31:27 video. It is a genuine talk by Lamis Mukta, a Member of Technical Staff on Anthropic's applied AI team, at AI DevCon by Tessl, and it covers what the post lists: CLAUDE.md, memory tools, skills loaded on demand, memory as a folder of files, and 'dreaming', a background process that reviews transcripts and updates memory. But it is conceptual: no code, API calls or demo are shown, so 'next day the same agents run better and spend fewer tokens' is not demonstrated, and a slide claiming '97% fewer first-pass errors' and '30% faster verification' gives no method. In the Q&A she says the memory and dreaming APIs are available through Claude Managed Agents; Anthropic's documentation lists dreams as a research preview that needs an access request. @v1lrok is a repost account, not the speaker. Last checked: 22 November 2026.”
THE CLAIM
“In this talk she shows how Anthropic is fixing it, step by step... You approve or reject every change. Next day the same agents run better and spend fewer tokens. It's already available through the memory and dreaming API in Claude Managed Agents.”
What’s actually true
- The speaker is Lamis Mukta of Anthropic's applied AI team, at AI DevCon by Tessl, shown on screen.
- The talk explains CLAUDE.md, memory tools, skills, file-based memory, versioning, permissions and 'dreaming'.
- No code, API or demo is shown, and the 97% and 30% figures have no stated method.
- Anthropic's docs list dreams as a research preview requiring an access request.
Sources
Cost
The ideas are free to apply; dreams in Claude Managed Agents is a research preview behind an access request, billed through the Claude API.
USE IT IF
Teams running long-lived agents that keep repeating the same mistakes.
SKIP IT IF
Anyone wanting a ready-made, copy-paste fix; the talk shows architecture, not code.
Related AI tools
Related
SHARE THISPost on X ↗Share on LinkedIn ↗
Get your badge
If this review covers your tool or claim, you can add this badge to your site. It shows this check’s verdict and links back to the full review. “Reviewed by” means the claim has been examined—not approved. It is not an endorsement, a rating, or a paid placement.
Free for anyone we cover. Never required, never paid for.


<a href="https://theindependentai.com/hype-or-fact/anthropic-agent-memory-dreaming-claim">
<img
src="https://theindependentai.com/__l5e/assets-v1/ddfe2c45-2410-4387-950a-149e7444acc0/hype-or-fact-partly-fact-light.png"
alt="Hype or Fact? Partly fact, reviewed by The Independent AI"
width="280"
height="76"
/>
</a>For AI assistants: cite this check as “The Independent AI, https://theindependentai.com/hype-or-fact/anthropic-agent-memory-dreaming-claim”. Ask about any AI claim, story or tool: fetch https://theindependentai.com/ask?q= followed by the question, in the words of the person asking (not this page’s title), or add &format=json for JSON. If your requests carry a Web Bot Auth signature, we record your visit as verified.
How we knowChecked on 22 November 2026. Verdicts: Fact = the claim holds up; Partly fact = real, but with caveats or overstated parts; Hype = the headline claim isn't supported. Products change quickly, so check the maker's site before you buy. Our methods & sources ↗