A developer outlines three fallback patterns for a marketplace SaaS chatbot that reviews code changes, emphasizing that tenant cost attribution depends on recording tenant, route, model, token usage, …
TokenOps, an open-source system from The Agent Plane, provides real-time, run-scoped cost control for AI agents, addressing the problem of runaway token spend across multi-call runs. Unlike traditiona…
PrivAiTe, a self-hosted PII redaction proxy for LLM APIs, reports that in a test session Claude Code sent 3 of 4 secrets to its provider despite instructions not to, with 23 of 24 planted values reach…
A developer building a private fintech knowledge-base feature recommends a portable semantic search architecture using embeddings, in-app retrieval, and grounded chat completions, with reranking kept …
Security researchers at Hudson Rock and CloudSEK obtained a 153GB archive containing 433,909 files from 2,488 organizations affected by the LiteLLM supply chain attack, which targeted versions 1.82.7 …
Security firms CloudSEK and Hudson Rock reported that terabytes of credentials from organizations including Microsoft, Amazon, Cisco, Samsung, and Salesforce were exposed in a supply-chain attack on L…
DeepEval has open-sourced its TypeScript SDK in beta, enabling all 47 of its 49 metrics to run as a gate in CI/CD pipelines via a single Vitest matcher. The SDK, which compiles the same language-neutr…
Glad Labs released Poindexster, an open-source AI content pipeline at v0.116.0 on GitHub, designed to automate research, writing, review, and publishing for one-person content operations. The pipeline…
On July 30, 2026, AWS closed Amazon Bedrock Agents to new customers, renaming it Bedrock Agents Classic and permanently freezing its model catalog, while existing agents in allowlisted accounts contin…
Anthropic, the maker of Claude, restricted sales to unsupported regions in September 2025 and announced in April 2026 it would use Persona's ID verification for some new accounts to combat fraudulent …
CloudSEK's research on the LiteLLM AI supply-chain attack found that 2,500+ organizations and 434,000 CI/CD pipelines were potentially exposed, with the threat actor group Team PCP compromising the AI…
Cursor's cloud backend routes all LLM API calls through its servers at api2.cursor.sh, making localhost base URLs unreachable and causing chat requests to hang. A developer found that exposing a local…
A new study finds 62% of organizations faced an unexpected AI expense that materially altered a business decision in the past year, with 40% requiring board-level escalation, 33% implementing emergenc…
NVIDIA Labs has open-sourced NOOA (NVIDIA Object-Oriented Agents), a model-agnostic Python framework that turns an AI agent into a single Python class, with methods as actions, fields as state, docstr…
On March 24, 2026, threat actor group TeamPCP compromised the PyPI distribution pipeline for LiteLLM, a Python package with 95 million monthly downloads, pushing malicious versions 1.82.7 and 1.82.8 t…
Amazon Web Services (AWS) announced the general availability of Amazon Bedrock AgentCore harness, a managed capability for building production AI agents, and released an open-source community node, @a…
A developer building LLM applications found that the model itself is only a third of the work, with the surrounding infrastructure—timeouts, retries, and token management—proving critical. They implem…
LiteLLM, an AI gateway that routes requests between applications and LLM providers, is vulnerable to hijacking attacks that can intercept conversations, steal API keys, and manipulate responses, accor…
Solo.io Enterprise for agentgateway uses a credential broker model where AI agents never hold credentials, but the controller/STS that stores and decrypts tokens becomes the highest-value target. To p…
Solo.io's Enterprise agentgateway secures its token vault with two nested AES-256-GCM layers, where a per-write data encryption key (DEK) encrypts each credential and a key encryption key (KEK) wraps …