SmartBear releases BearQ agentic QA system
SmartBear right this moment introduced the discharge of BearQ, an agentic QA system that may uncover how purposes work to make sure their integrity.
In line with the corporate, BearQ, part of SmartBear’s Utility Integrity Core product suite, can autonomously discover how purposes work and adapt to any adjustments in code, testing solely what issues. The flexibility to adapt makes this answer an evolution fron brittle automation scripts or predefined take a look at instances.
“The AI-driven hole between the velocity of software program improvement and the power to check ensuing purposes is getting wider, creating an enormous potential for software failures that harm corporations, customers and industries,” stated Dan Faulkner, SmartBear CEO, within the announcement. “BearQ will likely be a vital autonomous exploration and testing answer for the business to shut these gaps and ship enhanced software integrity. SmartBear’s lengthy historical past of innovation in software program high quality and testing makes SmartBear positioned to deal with this disruption to the SDLC.”
A current SmartBear research, Closing the AI Software program High quality Hole, discovered that 90% of respondents are utilizing AI coding instruments, however 70% say software program high quality is struggling. Additional, greater than three-quarters of respondents are nervous that testing gained’t have the ability to sustain with sooner AI code improvement, 9 of 10 survey respondents have adopted AI coding instruments. Seven of 10 are involved that high quality is already struggling, 60% have skilled high quality points up to now 12 months, and 68% are nervous that testing will likely be a bottleneck for software program supply.
BearQ additionally consists of guardrails and contextual controls that enable builders to outline what brokers should validate, and the way exploration and testing align with enterprise targets.
To be taught extra about BearQ and apply for GA Early Entry, click on right here.
Chainguard proclaims open supply Agent Abilities
Chainguard right this moment introduced Chainguard Agent Abilities, a catalog of hardened AI agent expertise that’s repeatedly maintained. Agent Abilities, which routinely ingest expertise from open supply registries, permits builders to develop the use instances of their brokers with out extending their assault floor. These expertise are measured towards a safety and high quality ruleset, hardens them utilizing Chainguard reconciliation brokers, and publishes them with a whole audit path. That vetting is essential, as a result of with out permission controls or oversight, agent expertise have grow to be the newest goal of provide chain assaults.
AI agent expertise are small, modular instruction units that reach what an AI agent can do. Builders set up expertise so as to add capabilities comparable to browser automation, PDF processing, database entry, and code-generation workflows. agent expertise have grow to be the newest goal of widespread provide chain assaults.
“Container pictures confirmed us how rapidly software program artifacts can grow to be provide chain dangers as soon as they’re adopted and trusted at scale. AI agent expertise are rising alongside a fair sooner trajectory,” stated Dan Lorenc, CEO and Co-founder, Chainguard. “As AI brokers grow to be embedded within the software program improvement lifecycle, the talents that form their conduct grow to be a part of the availability chain itself. With Agent Abilities, Chainguard is bringing steady hardening and verifiable integrity to that layer, so organizations can construct with AI on a safe basis.”
Chainguard Agent Abilities is offered in beta. To be among the many first to strive Chainguard Agent Abilities, go to https://www.chainguard.dev/agent-skills.