DAILY DROP · NOV 13, 2026

Today's drop: 1 new tool, 2 listing updates

The directory lists 150 hand-checked tools. New today: Bumblebee, a free, open-source, read-only scanner from Perplexity that checks developer machines for compromised packages, editor and browser extensions and MCP server configs when a supply-chain advisory lands. Updated: Claude Code (v2.1.323, Code Mode) and Semgrep (Mandoline research preview). AI for IT & Cybersecurity grows to 61 tools; we retired one Design listing to stay at 150. Pricing only where the official page states it.

Previous drop: Nov 12, 2026 →

Get the weekly dropThe best new AI tools, one email a week. Free.

New in the directory 1

Browse all 150
Tool
Category
Price
Status
Link
Bumblebee (Perplexity)Free, open-source (Apache-2.0) read-only scanner from Perplexity that answers one incident-response question: when a supply-chain advisory names a bad package, extension or version, which developer machines have it right now? It inventories language packages (npm, pnpm, Yarn, Bun, PyPI, Go modules, RubyGems, Composer, Homebrew), VS Code, Cursor, Windsurf and VSCodium extensions, Chromium-family and Firefox browser extensions, MCP server configs (Cursor, Claude Code, Claude Desktop, Cline, Gemini CLI) and agent-skill lock files, writes NDJSON records and flags exact matches against an exposure catalog. It reads metadata only: it never runs npm, pip or install scripts and does not record the secrets in MCP config env blocks. One static Go binary for macOS and Linux (go install), run once or on a schedule with cron, launchd or MDM; Perplexity maintains exposure catalogs for recent npm supply-chain campaigns in the repo. It is not an EDR, and Codex TOML MCP configs are not parsed yet. About 5,100 GitHub stars; latest tagged release v0.1.2 (Jun 2026), with catalog updates through Oct 2026.
Price Free and open source (Apache-2.0)
FREE TIERchecked 11/13

Why this one

  • Bumblebee answers the first question in every npm or extension supply-chain incident: which of our machines actually have the bad version? SBOMs cover what you shipped and EDR covers what ran, but the answer usually sits in messy local state: lockfiles, installed package metadata, VS Code and Cursor extensions, browser extensions and the MCP servers people added to Claude Code or Cursor. Bumblebee reads that metadata, writes one NDJSON record per component and flags exact matches against an exposure catalog; Perplexity keeps catalogs for recent npm campaigns in the repo. It is read-only on purpose: it never runs npm, pip or install scripts, so the scan cannot trigger the postinstall payload it is looking for, and it does not record the secrets that sit in MCP config env blocks. It is one static Go binary for macOS and Linux (no Windows build), Apache-2.0, with about 5,100 GitHub stars. It is not new this week (the repo dates from May 2026 and the last tagged release is v0.1.2), but it is the most practical free tool we have seen for this job, and it fits a home lab: run bumblebee on your own laptop, then read what it found in your MCP configs and extensions.

Listing updates

Checked on Nov 13, 2026 against official pages and release notes.

  • Claude Code: v2.1.323 (Nov 12) adds Code Mode, an experimental opt-in where Claude writes JavaScript to call its tools instead of calling them one at a time; admins can turn it off for an organization. Plugins can now declare which Claude Code versions they support in plugin.json, and out-of-range plugins are not installed or loaded. It also warns at startup about Bash allow rules with a * in a variable's value, such as Bash(PORT=* npm run dev), and the VS Code extension now asks you to trust a folder before the first session in it.
  • Semgrep: Semgrep Mandoline (Nov 12) is a research preview of program-analysis tools built for AI agents: call graphs and cross-file dataflow (taint) analysis across 50 languages that an agent can query while it investigates code. Semgrep says it is 50x faster than its Pro engine for these agent-style analyses. Existing customers get it inside Semgrep Agentic Workflows and the Context Layer; the free Community Edition listing is unchanged.

Category changes

  • AI for IT & Cybersecurity grows to 61 tools with Bumblebee.
  • Design drops to 6: we retired Uizard to stay at the 150-tool cap. Prompt-to-mockup is now covered better by Figma (Figma Make), v0 and Lovable, which also produce working code. Its old page redirects to the Design category.

Checked, not listed

  • agentaudit (Manifold Security) and Vaikora Guard MCP (Data443, now MIT open source on PyPI): agent config and MCP policy tools we are still watching for traction.
  • Fixify, ControlUp Pulse AI Resolve and SuperIT: new AI help desk products for MSPs and IT teams; business sales, not something you can try in a home lab yet.
  • Codex 0.170.0 alphas (Nov 12): pre-releases, no listing change. Ollama v0.40.17 and Wazuh 4.14.11 are still release candidates (rc0 and RC1), so their listings stay on v0.40.16 and 4.14.10.