Does your agent actually remember, or just sound like it? A 120-line open-weight test

·1 views

A friend asked me a question I couldn't answer about his own AI agent: does the thing answering me today actually remember being yesterday's, or is it just very good at sounding like it does? He owns the agent. He can't see its memory from the outside. And the people who sell the memory layer have every reason not to hand him a test that might fail. So I built him one.

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)

The Day Your AI Goes Rogue

The biggest enterprise AI risk in the next decade may not be a malicious AI. It will be a capable one pursuing the goal it was given — with too much access and nothing to stop it. For three years the enterprise conversation about AI safety has been about the answer. Is the output accurate. Is it biased.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

PostgreSQL Row-Level Security: Multi-Tenancy Without Leaking Tenant Data

Building a multi-tenant SaaS application is fraught with danger. In a traditional shared-database architecture, every single SQL query must remember to append a tenant filter: SELECT * FROM invoices WHERE tenant_id = 'acme_corp'; If a developer writes just one query and forgets WHERE tenant_id = :id: SELECT * FROM invoices WHERE id = :invoice_id; -- BUG! Tenant Acme Corp can suddenly view Tenant Globex's financial invoices! This is one of the most common and catastrophic data leak vulnerabilities in SaaS platforms. Instead of relying on fragile application-level WHERE clauses, PostgreSQL provi

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Ghost CMS Mass Disclosure — 6 CVEs Including CVSS 8.1 Staff Session Bypass

A Ghost staff user with valid credentials can log into any other staff account using only the target's password — bypassing 2FA entirely. CVE-2026-103283 (CVSS 8.1) leads a cluster of six vulnerabilities in Ghost CMS disclosed October 1. Here's the full picture: CVE-2026-103283 (8.1) — Staff session bypass. Any authenticated staff user can impersonate any other staff member with just their password. 2FA does not protect against this.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Why Data Reconciliation Matters in AWS Glue: Finding Missing Records with Anti Joins

A pipeline can be green and still have a data problem. This is one of the things I learned while working with data pipelines. A Glue job can complete successfully, files can be written to S3, and there may be no obvious error in the logs. But that doesn't necessarily mean that every record made it from the source to the target. That's where data reconciliation comes in.

0 comments Read more at DEV Community