// edition · 2026-04-08
32 stories on AI dev tools, agents, and the coding stack — curated from the day's RSS haul by Agentic Dev's pipeline.
Anthropic's Claude API tool use feature works by having Claude return JSON specifying which functions to call and with what arguments, rather than calling functions directly; users execute these functions and return results to Claude in an iterative loop.
An audit of 50 open-source MCP servers found 43% contained command injection vulnerabilities. The article outlines 22 security checks to prevent attacks, including avoiding shell string interpolation, eval/exec usage, and path traversal in servers that mediate between language models and producti...
MemCTX is an open-source MIT-licensed tool that maintains session memory for Claude Code by storing sessions in SQLite, auto-generating summaries via Claude API, and injecting relevant history into new sessions through a dashboard interface.
Waymark is an MCP server that intercepts file system and bash operations from Claude Code before execution, allowing users to set policies, log actions to SQLite, approve or reject operations via a web dashboard, and rollback changes.
A developer replaced their four AI agents' local model with Claude Opus accessed through a proxy wrapper, upgrading output quality while saving approximately $450/month in API costs through an existing Claude Max subscription at 200K tokens/day usage.
A developer launched SimplyLouie, a Claude API-compatible endpoint priced at $2/month flat rate, as an alternative to per-token billing models. The service uses Anthropic's official ANTHROPIC_BASE_URL mechanism to route API calls.
Vercel added an AI skill for coding agents and new CLI commands to manage Microfrontends, allowing users to create and manage microfrontend groups through natural language prompts or terminal commands without using the dashboard.
A developer purchased seven RTX 5090 GPUs for self-hosting an AI image and video generation service instead of renting cloud GPUs, citing AWS costs of $17,250–$45,120 monthly versus lower self-hosting expenses for a free-tier service generating 35,000 images and 4,000 videos daily.
Hybrid identity fraud using AI-generated faces is compromising biometric verification systems by creating synthetic IDs and liveness videos that match too perfectly, forcing developers to shift from simple facial matching to forensic analysis that detects shared synthetic origins through mathemat...
A tutorial demonstrates how to deploy self-hosted cloud infrastructure on a single server using Nextcloud, Collabora, and open-source tools, enabling real-time document collaboration, integration of four AI models through one interface, calendar sync, custom email, and automated backups.
Aria Networks announced a "Network that Thinks" initiative focused on optimizing Model Flop Utilization (MFU), a metric measuring datacenter hardware efficiency in AI clusters. The company argues that network infrastructure optimization directly affects token efficiency and cost-per-token in AI s...
Vercel added a query builder to Observability Plus that lets Pro and Enterprise teams create custom queries on workflow runs and steps, with filtering and grouping by environment, project, workflow, and step.
A developer released Octopal, an open-source app that lets users manage multiple Claude AI agents organized by project folder, with agents communicating via chat interface and an orchestrator routing tasks between them.
A developer built kerf-cli, a local CLI tool that analyzes Claude Code's session logs stored in SQLite, enabling detailed cost tracking and optimization recommendations that Claude Code's native `/cost` command does not provide.
Anthropic released a system card for Claude Mythos Preview, a new model variant, documenting its capabilities and characteristics.
Anthropic launched Project Glasswing and Claude Mythos Preview, an AI model that discovered zero-day vulnerabilities in OpenBSD, FFmpeg, Linux, and browsers that human audits and automated fuzzers missed for years. The company chose not to release the model publicly due to cybersecurity risks.
Anthropic released Claude Mythos Preview with accompanying security documentation and initiated Project Glasswing to address software security in the AI era.
Z.ai released GLM-5.1, a 754-billion-parameter open-source AI model available under MIT license. The model demonstrated the ability to generate SVG graphics with CSS animations and to debug and fix code issues when given follow-up instructions.
Z.ai's GLM 5.1 model is now available on Vercel's AI Gateway, supporting long-horizon autonomous tasks including multi-step coding workflows, conversation, creative writing, and document generation.
Vercel made Fast Mode available for Claude Opus 4.6 on AI Gateway, offering 2.5x faster output token speeds at 6x standard pricing. The experimental feature is designed for human-in-the-loop workflows and coding tasks.
An engineer argues that while AI can help polish technical writing, relying on it to generate content about unfamiliar topics produces superficially well-written but substantively empty work. Authentic technical writing requires personal experience and context from real debugging and production i...
Anthropic designed Claude using constitutional AI principles prioritizing safety over capability, resulting in a system that refuses requests more frequently and produces more conservative outputs. The approach creates a trade-off where increased safety constraints limit creative tasks like story...
Anthropic restricted third-party Claude tools like OpenClaw to a separate pay-as-you-go plan starting April 4, 2026, and issued DMCA takedowns against repositories hosting leaked Claude Code source code. The moves drew criticism as contradictory given Anthropic's prior settlement ($1.5B in Septem...
Anthropic released a preview of its Mythos AI model for use by select companies in defensive cybersecurity work.
Anthropic restricted access to Claude Mythos, a new AI model demonstrating advanced autonomous exploit development abilities, through Project Glasswing to let industry partners patch vulnerabilities before broader capability proliferation. Mythos has already identified thousands of high-severity ...
Software pioneers Kent Beck and Martin Fowler discussed at the Pragmatic Summit how AI adoption cycles resemble previous tech disruptions, warning that misaligned incentives and poor performance metrics may repeat patterns seen with Agile, while emphasizing test-driven development's continued rel...
Anthropic released Claude Mythos Preview, a new frontier AI model, to about 50 select partners including Amazon, Apple, and Microsoft through Project Glasswing for defensive cybersecurity work. The model scores 83.1% on vulnerability analysis benchmarks, compared to 66.6% for the previous flagshi...
Simon Willison's research confirms that SQLite WAL mode functions correctly across separate Docker containers sharing the same volume, as containers on the same host share the same shared memory required for WAL coordination.
The New Yorker published an 18-month investigation finding a discrepancy between Sam Altman's public statements on AI safety and OpenAI's actual spending and practices in the area.
Anthropic launched Project Glasswing on April 7, 2026, a cybersecurity partnership with 11 major tech companies using an AI model called Claude Mythos Preview to identify vulnerabilities in software. The model discovered thousands of zero-day vulnerabilities including a 27-year-old bug in OpenBSD...
Anthropic launched Project Glasswing, a security initiative aimed at securing critical software in the context of AI development and deployment.
Arcee, a 26-person U.S. startup, developed a high-performing open source large language model that is gaining adoption among OpenClaw users.
Free daily email. Every story you just read, in your inbox at 6 AM ET.