Meta unveils Muse models, Anthropic replaces Git, and Goodfire monitors agents - AI Programming Weekly #72

October 13, 2026 - AI Programming Weekly
Hi there,

Welcome to this week's edition of AI Programming Weekly.

Quilty brings screenplay analysis tools into Claude and ChatGPT, while Edge Delta releases the open source AI SRE Arena benchmark for testing AI site reliability engineering tools. Several companies are also extending AI agent access through the Model Context Protocol, including GreenCore Solutions, Ploti, Precisely and Incognia.

Orgvue adds a WebMCP interface for workforce design, and Trellist reports a 30% reduction in customer service calls in a logistics deployment. This edition also covers REA and ArtCraft, whose AI assisted software recreation projects have raised questions about reliability, compatibility and intellectual property rights.
Quilty brings screenplay analysis to Claude and ChatGPT
Quilty makes its screenplay coverage available in Claude and ChatGPT, including scores, story notes, character feedback, and market research. The platform also compares draft versions and tracks score changes across four weighted criteria. Read more
IQuest Research releases IQuest-Q1 open weights
IQuest Research releases public model weights and technical materials for IQuest-Q1, a 320 billion parameter model for coding, software engineering and multi step agent tasks. The company reports evaluations spanning repository code generation, terminal use and professional workflows. Read more
Edge Delta launches AI SRE benchmark after detecting 18 of 21 incidents
Edge Delta introduced its telemetry native AI SRE platform and the open source AI SRE Arena benchmark for assessing incident response systems. In an initial test involving 21 Kubernetes production incidents, the company reports its system detected 18 incidents autonomously. Read more
ActivTrak launches MCP integration for AI work intelligence
ActivTrak introduces ActivTrak MCP, connecting behavioral work data with CRM, HRIS and ticketing data for AI analysis of workforce capacity, productivity and performance. Early access is available through Claude and ChatGPT, with Copilot and Gemini support planned. Read more
Waters launches AI Link for Anthropic connected LC MS workflows
Waters launches AI Link, integrating Anthropic's Model Context Protocol with waters_connect software for AI directed LC MS workflows. The company also joins Anthropic's Model Hardware Standard research program, with general availability of AI Link planned for November 2026. Read more
Meticulous raises $15 million Series A for autonomous AI code testing
Meticulous raises $15 million in a Series A led by Chemistry, with participation from Menlo Ventures and angel investors. Its platform generates and maintains software tests to help developers and AI coding agents assess code changes before they are merged. Read more
Anthropic builds Scion to replace Git for its monorepo
Anthropic says it migrated its monorepo from Git to its in house version control system, Scion, in five months. The company reports faster checkouts, lower CI CPU use and support for a substantial increase in daily code merges. Read more
Anthropic launches free AI vulnerability scanner for open source projects
Anthropic offers eligible open source projects periodic vulnerability scans powered by its AI models, including Claude Mythos. The reports can identify suspected flaws and propose patches, but maintainers must review findings because they receive no human verification. Read more
Orgvue adds WebMCP interface for AI agents
Orgvue introduces a WebMCP interface that allows customers to use AI agents for workforce design tasks. The platform requires users to review and approve actions, with organizational changes recorded in auditable, reversible decision records. An early access trial is underway, and general availability is planned for January 2027. Read more
Incognia launches AI agent detection for financial institutions
Incognia introduces AI Agent Detection, a tool for financial institutions to identify AI agent activity and evaluate actions performed for customers. The company also adds Web Behavioral Biometrics and an MCP Server for read only, AI assisted fraud investigations. Read more
Trellist cuts logistics operator's customer service calls by 30% with AI
Trellist reports that its AI project for an unnamed national logistics, transportation, and storage operator reduced customer service call volume by 30%. The project connected operational data and introduced AI tools for development, quality assurance, and customer service with human review. Read more
Linux Foundation report finds 94% of European organisations use or pilot AI coding tools
A Linux Foundation report finds that 94% of European organisations use or pilot generative AI coding tools. The study also highlights digital sovereignty concerns, gaps in open source governance, and increased investment in dependency management. Read more
Meta introduces Muse Wonder AI models
Meta introduces Muse Wonder and Muse Wonder Full, two AI models for complex personal and professional agent tasks. The models are available through the Meta Model API and Muse Code, with Muse Wonder Full limited to Muse Code subscribers. Read more
Movable Ink launches headless marketing via MCP Server
Movable Ink launches headless marketing capabilities through an MCP Server, allowing marketers to manage campaigns in its platform or through external AI tools including Claude and ChatGPT. The server provides access to more than 50 APIs across eight areas, supporting tasks such as analytics, brand setup, subject line creation and creative development. Read more
Apollo GraphQL introduces governed access for enterprise AI agents
Apollo GraphQL introduces GraphOS Agent Services, which provides AI agents with controlled, field level access to enterprise APIs and systems. The service includes identity management, policy controls and audit logs, with Intuit piloting it. Read more
Validity adds MCP connector to Engage for email intelligence
Validity adds a Model Context Protocol connector to Engage, allowing enterprise marketers to access email and competitive intelligence through their existing AI tools. Engage also adds pre send checks for layouts, links, accessibility, branding and compliance. Read more
Harness acquires select Augment Code assets for software factory agent
Harness acquired selected assets from Augment Code, including Cosmos, Auggie CLI, and the Code Context Engine. Cosmos is now offered as the Harness Cosmos Software Factory Agent, which automates engineering tasks from initial ideas through merge ready code. Read more
Mashreq pilots IBM Bob for agentic AI integration engineering
Mashreq is piloting IBM Bob to support integration engineering, software delivery and troubleshooting. The UAE bank reports early time reductions of 60% to 70% for integration design and troubleshooting, and is also evaluating IBM Context Forge and an on premises deployment. Read more
TestMu AI Named a Leader in Gartner's 2026 Magic Quadrant
TestMu AI says Gartner positioned the company as a Leader in its 2026 Magic Quadrant for Agentic Software Quality Assurance Platforms. The Noida based company provides agentic testing, quality engineering, test orchestration and cloud execution tools for enterprise teams. Read more
Abstracta introduces Quality Intelligence for AI era software releases
Abstracta introduces Quality Intelligence, a framework for assessing change risk and release readiness as AI accelerates software development. The company applies the framework through Abstracta Intelligence, an enterprise platform offering AI agents and dashboards for organizations operating critical software. Read more
Inside Real Estate launches BoldTrail MCP for ChatGPT and Claude
Inside Real Estate launches BoldTrail MCP, connecting its real estate platform to AI assistants including ChatGPT and Claude through the Model Context Protocol. The integration lets agents, teams and brokerages manage contacts, tasks, appointments and communications using natural language. Read more
GreenCore Solutions releases gtinfetch.ai GTIN resolver for AI agents
GreenCore Solutions releases gtinfetch.ai, a GTIN resolver that lets AI agents identify up to 84 million product barcodes through a single call. The service uses the Model Context Protocol, runs on Microsoft Azure and Cloudflare, and supports USDC and Open USD payments on Base. Read more
Goodfire launches AI agent monitors on Baseten
Goodfire launches monitoring probes for AI agents on Baseten that analyze internal model signals for risks including hacking, weapons misuse and reward hacking. The company reports that tests on Kimi K3 detected 93% of malicious hacking sessions at a lower cost than using a separate AI model for monitoring. Read more
Epic Games pushes AI first development for delayed Disney shooter
Epic Games reportedly told teams developing an unannounced Disney shooter featuring Marvel and Star Wars characters to use Anthropic's Claude models for art and coding. The project has reportedly been delayed from its previously expected March 2027 release amid slow progress and staffing shortages. Read more
Ploti launches AI data layer for real estate records in 40 states
Ploti launches a data layer that gives AI agents access to source traced local government records, including zoning, permits, ownership, flood and utility data. The service covers 57,818 datasets across 40 US states and supports assistants using the Model Context Protocol. Read more
Voya Financial Selects UnqorkAI for Retirement Workflows
Voya Financial selects UnqorkAI to support AI enabled retirement services, streamline onboarding and connect customer support workflows. The agreement expands a partnership between the companies that began in 2019. Read more
Gremlin makes Foresight AI generally available
Gremlin makes Foresight AI generally available following a beta period. The platform uses Failure Atlas to identify reliability risks, recommend or apply infrastructure and configuration code fixes, then retest systems. Read more
Gusto redesigns Gusto Pro with AI workflows for accountants
Gusto introduces a redesigned Gusto Pro with AI assisted payroll workflows, multi client tools, automated reporting, and Gmail and Outlook connectors. The company also adds a multi client Model Context Protocol and expands its partner program. Read more
Speakeasy and Tailscale add private networking for enterprise MCP
Speakeasy and Tailscale integrate private networking for enterprise MCP gateways. The integration keeps MCP traffic on customer Tailscale networks while applying identity, governance, policy and auditing controls through the Speakeasy AI Control Plane. Read more
Chainguard launches Harness for AI vulnerability verification
Chainguard launches Harness, a security tool that uses frontier AI models, including Anthropic's Claude Mythos 5.1, to identify exploitable vulnerabilities in live application builds. The tool reproduces findings in isolated microVMs and provides staff reviewed codebase specific fixes. Read more
Precisely adds MCP servers for AI agents
Precisely adds Model Context Protocol servers that let compatible AI agents connect with its data management and customer communications tools without custom integrations. The servers support data cleansing, address verification, SAP automation, master data search and customer communications. Read more
Market Logic Software makes DeepSights MCP generally available
Market Logic Software makes DeepSights MCP generally available for customers. The connection allows assistants including Microsoft Copilot, ChatGPT, Claude and Glean to access cited market intelligence under existing enterprise permissions. Read more
REA and ArtCraft create AI-built alternatives to Adobe and Microsoft software
REA, an AI-assisted reverse engineering toolkit, and ArtCraft, a project by developer Brandon Thomas, are being used to create applications modeled on Adobe and Microsoft software. The efforts have drawn attention alongside questions about reliability, compatibility and intellectual property rights. Read more
Revenue Growth Agent launches MCP connector for ChatGPT and Claude
Revenue Growth Agent launches a Model Context Protocol connector for supported ChatGPT and Claude environments. It provides company specific sales intelligence and agents for meeting preparation, deal risk analysis, proposals and CRM workflows. Read more

We hope you enjoyed this article

Consider subscribing to one of our newsletters like AI Programming Weekly or Daily AI Brief.

Also, consider following us on social media:

Software Engineering AI Rene van Pelt AI Brief AI Brief (X)
Free newsletter

AI Programming Weekly

Weekly news about AI tools for software engineers, AI enabled IDE's and much more.

AI Token Webinar