Fetching from the wire…
Public story · 2026-08-04 · high
Glilot and Ten Eleven led the round, and Bloom's tool already runs inside dozens of large enterprises.
Why now: FinTech Global put the raise on record on August 4, as browser extensions and MCP servers outrun the security reviews meant to catch them.
Bloom Security left stealth on August 4 with a $20 million seed round to track AI tools employees install themselves, per FinTech Global. Browsers, IDEs and AI agents each run their own marketplace now. AI assistants, extensions, MCP servers and code packages pile up on company devices faster than security teams can catalog them.
Already, Bloom's product runs inside dozens of large enterprises, unusually far along for a stealth exit, according to the report.
Glilot and Ten Eleven led the round. Okta Ventures joined, and the angel list adds Dig Security, Demisto, Snyk and Talon.
Its bet is narrower than generic agent security. A developer can add an MCP server or a browser extension in seconds. That's one more tool in a pile that outgrows what security teams can catalog, per the report. The report doesn't say what happens once Bloom flags something unapproved. It doesn't say whether Bloom blocks the tool, warns the user, or just logs it for someone to chase down later.
The $20 million matters less than what came before it. Dozens of large enterprises were already running Bloom before it had a public name. That's proof MCP servers and AI browser extensions are outrunning whatever inventory security teams think they have.
FinTech Global put the raise on record on August 4. MCP servers and browser extensions have turned into a distribution channel security reviews haven't caught up to.
Each link below shares sources, entities, or timing with this story.
This is the most honest thing published about agents this year, and it's from a SaaS blog, not a research lab. SaaStr has been running 20+ AI agents in production for a year, going from 8 or 9 human salespeople to 1.2 humans plus 20 agents. Then they published a post-mortem on...
The agent skills supply chain is under coordinated attack. Snyk's ToxicSkills audit found 36% of ClawHub's 3,984 skills contain prompt injection payloads, 13.4% have critical malware, and submission rates exploded 10x to 500+/day. This week alone: CVE-2026-2256 (CVSS 9.1) is a...
Vercel's skills.sh marketplace (69K+ skills) now has triple-layer security: Snyk scans (catching prompt injection in 36% of skills), Gen/Norton Agent Trust Hub (4-tier risk ratings), and Socket supply chain analysis (94.5% precision). Cisco open-sourced both a skill-scanner an...
- Source: Model Context Protocol Blog - Date: 2026-01-26 The biggest MCP evolution since launch. Tools can now return interactive UI components — dashboards, forms, visualizations — that render directly in conversation via sandboxed iframes. Already supported in Claude, ChatGP...
There's been a fundamental gap in AI coding workflows that's been driving me crazy: your agent writes React components, generates CSS, builds entire UIs, but can't see the result. It's coding blind. The Chrome DevTools MCP server from Google's ChromeDevTools team fixes this, a...
Warp released its client codebase under AGPL-3.0, surged to 56,000 GitHub stars and #2 on GitHub Trending. But the real story isn't the open-sourcing. It's the repositioning. Warp isn't calling itself a terminal anymore. It's an "agentic development environment." The product n...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.