Find published research by company, product, platform or technology.
Showing 101–120 of 219 dossiers
MLflow CrewAI autologging silently logged API keys; 3.17.0 patches it, but previously exported traces and keys still need remediation.
WebMCP is no longer a Chrome-only browser experiment: Microsoft Edge now has its own active origin trial, while ChatGPT’s built-in browser and WordPress Playground show agent-client and platform implementation paths.
Bitwarden is changing which client binaries it distributes through app stores, not retroactively relicensing GPL code. Self-hosters and downstream redistributors should distinguish current compatibility from future feature divergence.
Beam makes a large agent-focused model commercially reusable without a bespoke licence, but the open weights are not a small local download.
A new GitHub ruleset can stop unapproved dependency licences at merge time, and agents can check packages before recommending them.
The October 15 Gateway/Workers integration turns a secure web gateway into an application-specific programmable proxy. Teams gain custom request/response handling, but must account for inspection order, WebSocket exclusions and standard Workers billing.
The October 15 Forgejo 17.0 release strengthens self-hosted Git and CI workflows, but webhook receivers need a migration because the new workflow-run events change headers and payload semantics. The release is non-LTS.
CVE-2026-105445 moved personal access tokens across Docker Hub identities without rotating their secrets. Affected users were notified and tokens deactivated after the September 30 fix, but the two-year exposure illustrates a container supply-chain trust boundary.
After Chrome 153's signed-JSON migration, Chrome 154 brings Android and third-party origin-trial support and changes the issuance request's Sec-Fetch-Dest value. Chrome 156 will reject provider tokens whose email claim does not exactly match the submitted address.
RC2 is scheduled for November 12, GA for November 19. New concurrent REPACK, parallel autovacuum, replication behaviour and planner changes remain pre-release; developers should test extensions, query plans and upgrades rather than deploying the RC to production.
Europe's new large-model contender is available to test through Mistral Studio, but its open-weight promise, pricing and benchmark claims need careful qualification.
The important part of pg_vault_tde's 1.7.2 release is the operational migration: v4 rows can still be read after upgrade, but UPDATE can crash until they are rewritten.
The certificate lifetime change is not a new CA or endpoint. It is a renewal-timing deadline for operators whose cron jobs assume 90-day certificates.
PostgreSQL operators gain per-statement and per-transaction estimated-cost limits across versions 14–18, useful for runaway reports and ORMs. The guard is disabled by default and can be bypassed by users allowed to change planner cost parameters.
The October Nuxt release lays groundwork for server-engine portability and addresses TypeScript scaling problems in large route graphs without claiming Nitro has already been replaced.
The sharpest operational trap is CVE-2026-91768: PHP-FPM’s listen.allowed_clients can treat an allowed IPv6 address as an entire /96 prefix. Teams relying on that ACL should patch rather than assuming the configured address restriction is exact.
MiMo-V2.6 is more useful than another benchmark launch because builders get both capable multimodal weights and a rare view into the reinforcement-learning machinery that produced them: code, environments, run costs and even failure notes from the training cluster.
GitHub-hosted Actions jobs that use `ubuntu-latest` are about to change operating-system generation without a YAML edit; teams can test on `ubuntu-26.04` now or pin 24.04 while they migrate.
The change is not about where database rows live; Cloud SQL already has regional instance placement. It changes where API control traffic is processed, reducing dependence on global frontend infrastructure and making data-in-transit boundaries easier to align with sovereignty requirements.
SwarmLLM does not route whole prompts to separate machines; it pipelines one model across browser tabs. A MacBook and iPhone can jointly hold Qwen 3.8 27B even when neither device can hold the full 15GB quantized model alone, with no inference server in the loop.