Mabl declares updates to Agentic Testing Teammate
The Agentic Testing Teammate works alongside human testers to make the method extra environment friendly. New updates embody AI vectorizations and take a look at semantic search, enhancements to check protection, and enhancements to the MCP Server that allow testers to do various duties immediately inside their IDE, together with Check Affect Evaluation, clever take a look at creation, and failure suggestions.
“This new work is constructed on the concept an agent can turn into an integral a part of your testing group,” stated Dan Belcher, co-founder of mabl. “In contrast to scripting frameworks and general-purpose giant language fashions, mabl builds deep information about your utility over time and makes use of that information to make it–and your group–more practical.”
Couchbase 8.0 provides three new vector indexing and retrieval capabilities
These new capabilities are designed to assist various vector workloads that facilitate real-time AI functions.
Hyperscale Vector Index relies on the DiskANN nearest-neighbor search algorithm and allows operation throughout partitioned disks for distributed processing. Composite Vector Index helps pre-filtered queries that may scope the precise vector being sought. Search Vector Index helps hybrid searches containing vectors, lexical search, and structured question standards in a single SQL++ request.
Anthropic expands reminiscence to all paid Claude customers
Anthropic introduced that the current reminiscence characteristic in Claude is being rolled out to Professional and Max plan customers, making it accessible to all paid customers now.
Reminiscence was initially introduced in early September, however was solely accessible to Crew and Enterprise customers to start with.
Reminiscence permits Claude to recollect your initiatives and preferences so that you simply don’t have to re-explain vital context throughout classes. “Nice work builds over time. With reminiscence, every dialog with Claude improves the following,” Anthropic wrote in its preliminary announcement.
Harness brings vibe coding to database migration with new AI-Powered Database Migration Authoring characteristic
Harness is on a mission to make it simpler for builders to do database migrations with its new AI-Powered Database Migration Authoring characteristic. This new functionality permits customers to explain schema modifications in pure language to obtain a production-ready migration.
For instance, a developer may ask “Create a desk named animals with columns for genus_species and common_name. Then add a associated desk named birds that tracks unladen airspeed and correct title. Add rows for Captain Canary, African swallow, and European swallow.”
Harness’ platform would then analyze the present schema and insurance policies, generate a backward-compatible migration, validate the change for security and compliance, commit it to Git for testing, and create rollback migrations.
Pink Hat Developer Lightspeed brings AI help to Pink Hat’s Developer Hub and migration toolkit
Pink Hat Developer Lightspeed has been built-in into each the Pink Hat Developer Hub and the migration toolkit for functions (MTA).
Within the Pink Hat Developer Hub, it acts as an assistant to hurry up non-coding duties, like exploring utility design approaches, writing documentation, producing take a look at plans, and troubleshooting functions.
Within the migration toolkit, Pink Hat Developer Lightspeed automates supply code refactoring inside the IDE. It leverages MTA’s static code evaluation to grasp migration points and how one can repair them, and in addition improves over time by studying what made previous modifications profitable.
GitKraken releases Insights to assist corporations measure ROI of AI
GitKraken, an organization that focuses on enhancing the developer expertise, introduced the launch of GitKraken Insights to supply corporations with higher insights into AI’s impression on developer productiveness.
In response to the corporate, whereas many engineering groups have adopted AI at this level, it’s nonetheless a problem to show AI’s ROI. GitKraken additionally believes that conventional engineering metrics weren’t designed for the AI period.
“DORA metrics can let you know if deployment frequency modified, however they’ll’t let you know why,” the corporate wrote in a weblog submit. “Pull request counts may go up, however is that as a result of builders are extra productive, or as a result of AI is producing code that requires extra overview cycles?”
GitKraken Insights brings collectively a number of completely different metrics—DORA metrics, code high quality evaluation, technical debt monitoring, AI impression measurement, and developer expertise indicators—to color an image of what’s occurring inside the growth lifecycle.
MariaDB unifies transactional, analytical, and vector databases in MariaDB Enterprise Platform 2026 launch
MariaDB’s Enterprise Platform 2026 launch was introduced this week, with the promise that it’ll act as “the definitive database platform for constructing next-generation clever functions.”
To assist agentic AI, the corporate added native RAG for grounding LLMs with context from MariaDB while not having embeddings, vector shops, or retrieval pipelines. The corporate additionally added ready-to-use brokers inside the platform, together with a developer copilot that connects to the database and might reply to pure language queries, and a DBA copilot that may handle duties like efficiency tuning and debugging.
Moreover, the corporate added an built-in MCP server in order that brokers can work together with MariaDB databases. The MCP interface in MariaDB permits customers to combine vector search, LLMs, and commonplace SQL operations, and permits brokers to launch serverless databases within the cloud.
Spotify Portal now usually accessible and filled with options for enhancing dev expertise
Spotify Portal for Backstage supplies builders with a ready-to-use model of Backstage, its open supply resolution for constructing inside developer portals (IDPs).
AiKA, which is an AI assistant for Portal, can now hook up with third-party MCP servers and set off actions in Portal. AiKA itself additionally capabilities as an MCP server, permitting builders to attach it as much as instruments like Cursor or Copilot and entry Portal knowledge.
“The overall availability of Spotify Portal marks a pivotal second in how organizations construct, measure, and optimize developer expertise. What started as an inside instrument for Spotify engineers is now a fully-fledged platform for enterprises, combining the reliability of Backstage, the perception of Confidence, and the velocity of AI-driven workflows,” Spotify wrote.
Sonar declares new resolution to optimize coaching datasets for coding LLMs
Sonar, an organization that focuses on code high quality, introduced a brand new resolution that may enhance how LLMs are educated for coding functions.
In response to the corporate, LLMs which are used to assist with software program growth are sometimes educated on publicly accessible, open supply code containing safety points and bugs, which turn into amplified all through the coaching course of. “Even a small quantity of flawed knowledge can degrade fashions of any dimension, disproportionately degrading their output,” Sonar wrote in an announcement.
SonarSweep (now in early entry) goals to mitigate these points by guaranteeing that fashions are studying from high-quality, safe examples.
It really works by figuring out and fixing code high quality and safety points within the coaching knowledge itself. After analyzing the dataset, it applies a strict filtering course of to take away low-quality code whereas additionally balancing the up to date dataset to make sure it can nonetheless provide various and consultant studying.
