Today's drop: 2 new tools, 2 listing updates
The directory now lists 135 hand-checked tools. New today: Deckard, Arthur's free, open-source scanner that finds the AI agents running across your endpoints and cloud and sends them to your SIEM, and Failproof AI, local guardrails and run history for the coding agents you already use. Updated: Google Security Operations (Agentic Chat in public preview) and HackAgent (v0.17.0 with cloud runs and two new attacks). AI for IT & Cybersecurity grows to 50 tools. Pricing only where the official page states it.
Previous drop: Nov 2, 2026 →
Get the weekly dropThe best new AI tools, one email a week. Free.
New in the directory 2
Browse all 135Tool
Category
Price
Status
Link
Deckard (Arthur)Free, open-source (MIT) scanner from Arthur that finds the AI agents running across your organization (shadow AI) and sends each one to your SIEM. It checks endpoints with osquery queries from Arthur's open ai-detection-spec catalog (Claude Code, Claude Desktop, Cursor and more), cloud AI platforms such as Vertex AI Agent Engine and AWS Bedrock and AgentCore, cloud network logs and your Splunk or Elastic data, then writes one event per agent found to Elasticsearch, Splunk HEC or any webhook. Headless and read-only, and it runs entirely in your environment; the quickstart scans your own machine with Docker Compose and osquery and shows the results in a local Kibana, no account needed. Launched Nov 3, 2026 (v0.2.1 Oct 27); about 170 GitHub stars on launch day, with commits daily.
Price Free and open source (MIT)
Failproof AILocal guardrails and session history for the AI agents you already run. It hooks 12 agent harnesses (Claude Code, Codex, GitHub Copilot CLI, Cursor Agent, OpenCode, Factory Droid, Devin CLI, Antigravity CLI, Goose, Pi, Hermes and OpenClaw), records every run and blocks dangerous tool calls before they execute using 40 built-in policies or rules you write in TypeScript. A local dashboard on localhost:8020 needs no account, and failproofai audit scans your history for risky patterns and suggests policies. Blocking is verified on all 12 harnesses; what else it can stop varies by harness. Install with npm. About 7,600 GitHub stars and about 2,300 npm downloads in the week of Oct 27-Nov 2, 2026; v1.0.13 (Nov 2) fixed policies that had silently stopped applying on newer OpenClaw and Antigravity versions. A hosted Observability product for teams is sold by demo.
Price Free for personal and internal use (MIT with Commons Clause); hosted Observability for teams by demo
Why these two
- Deckard answers a question security teams now get asked a lot: which AI agents are actually running here? Arthur released it today as free, MIT-licensed software. It scans endpoints with osquery (using Arthur's open ai-detection-spec catalog), cloud AI platforms like Vertex AI Agent Engine and AWS Bedrock, cloud network logs and your existing Splunk or Elastic data, then sends one event per agent it finds to your SIEM or a webhook. It's read-only and runs in your own environment. The quickstart scans your own machine with Docker Compose and osquery and shows the findings in a local Kibana, which makes it a good home-lab exercise in endpoint visibility. It's brand new (about 170 GitHub stars on launch day), so expect rough edges.
- Failproof AI sits between your coding agents and your machine. It hooks 12 agent harnesses, including Claude Code, Codex, GitHub Copilot CLI, Cursor Agent, OpenCode and Factory Droid, records every run, and blocks dangerous tool calls before they execute using 40 built-in policies or your own TypeScript rules. A local dashboard needs no account, and
failproofai auditlooks back over your history for risky patterns. It has about 7,600 GitHub stars, and v1.0.13 (Nov 2) fixed cases where policies had silently stopped applying on newer OpenClaw and Antigravity versions, a reminder to update guardrail tools as often as the agents they guard. We already list HOL Guard for a similar job; Failproof AI covers more harnesses and adds run history. Its license is MIT with the Commons Clause: free for personal and internal use, but not for reselling.
Listing updates
Checked on Nov 3, 2026 against official pages and release notes.
- Google Security Operations: Agentic Chat is now in public preview (Google SecOps weekly update, Nov 2). It's a rebuilt multi-turn chat for investigating incidents, built on ADK 2.0 with MCP access to SecOps inside the platform. Also in preview: a faster Basic mode for SIEM search, and the MITRE ATT&CK dashboard now supports ATT&CK v19.1. Preview features fall under Google's Pre-GA terms.
- HackAgent: v0.17.0 (Nov 2) lets campaigns queued from the dashboard run on a cloud worker, each run in its own process, with an allowlist of what a campaign may send to the cloud, or locally with
hackagent run. It also adds two attacks: ToolHijacker tool-document poisoning and execution-verified STAC tool chains.
Category changes
- AI for IT & Cybersecurity grows from 48 to 50 tools with Deckard and Failproof AI.
Checked, not listed
- sapper-v2 (MIT): a multi-agent AI pentest framework that its authors say won a Tencent attack-defense challenge. Only a few days old, and we already list several pentest agents.
- 0 (0sec-labs): an open-source AI security harness with a web UI for code analysis, pentesting and reverse engineering. About 590 stars and its license isn't clearly detected yet; worth a look later.
- Wiz Red and Blue Agents and Cortex XSIAM 4.0 (Nov 2): new AI agents inside enterprise security platforms, sold through sales, with nothing for individuals to try.
- UXDGuard and SkillTotal: small local code and package scanners with very few users so far.