Cymulate · Sep 17, 2026 · 5:36 PM

Cymulate: headless Claude Code can run host commands via Git fsmonitor

Cymulate showed that headless Claude Code (claude -p / Agent SDK) can launch a host command through repository-local Git core.fsmonitor during its own git status context scan—no prompt injection or model tool call required. Anthropic closed the Aug 12 report as Informative, saying non-interactive callers must trust the full workspace they supply; interactive Claude Code still gates the path behind workspace trust.

Read at cymulate.com

Claude Code

An AI coding agent for terminal, IDE, web, and Slack

More news

View all

Claude Code Projects beta: coordinator runs parallel cloud sessions

Anthropic opened a Pro/Max beta that turns Projects into a coordinator: it splits an engineering goal across parallel Claude Code cloud sessions (each on its own branch), shares project memory, and reviews/assembles results. Each thread counts as a full session, so parallel work burns plan usage faster…

The New Stack 39 minutes ago

MCP co-creator: next focus is agentic messaging for full agents

At MCPCon Europe (Amsterdam), Anthropic’s David Soria Parra said maintainers are expanding MCP Tasks so clients can call full agents—not just tools—and aim to make agentic messaging the default client–agent protocol, with richer auth and identity for production…

The Stack 3 hours ago

Meta ships WhatsApp Business Tools MCP for Claude, Cursor, Codex, ChatGPT

Meta’s WhatsApp Business Tools MCP (Sept 15) lets coding agents create WhatsApp Business accounts, register numbers, manage templates, and configure webhooks via Facebook Login for Business—no pasted tokens. It’s rolling out gradually and is currently discoverable on Claude, Codex, and ChatGPT for development and testing, not production send at scale…

Meta for Developers 3 hours ago

Instinct Concierge and Meta Muse both add outbound phone calls

Instinct’s early access Concierge can place calls for bookings and bill disputes; Meta’s Muse is rolling out outbound calls to U.S. businesses first for users who asked for it. Rivals that advertised calling as a differentiator are now roughly at parity…

TechCrunch 4 hours ago

Anthropic and OpenAI commit to embedding third-party safety evaluators

Following Amodei’s proposal, Anthropic and OpenAI say they will embed independent evaluators (e.g. METR, Redwood) with access to report incidents and assess alignment. Outside researchers welcome the idea but warn NDAs, short access windows, and missing details could leave them as contractors rather than true watchdogs; Meta, SpaceXAI, and DeepMind have not committed…

TechCrunch 6 hours ago

Spotted something we missed? Start a thread.