Updated

Open-source local security layer for AI coding agents that audits past Claude Code, Codex and Cursor sessions for injected commands and leaked keys, and installs pre-tool-call hooks that block or question risky actions next time.

Pricing Free open source
Visit site

Follow Stroq to get an email when its listing changes, such as new pricing or features. One digest a day at most, and you can unsubscribe anytime.

Company Agybay Gibatollayev
Pricing model Open Source
Paid plans None listed
Category peers 25
01Key features
  1. 01

    `stroq` card summarizing 30 days of agent sessions: commands that came from text the agent read and keys that left the machine

  2. 02

    `stroq sent` traces where each known credential went, printing key names, never values

  3. 03

    `stroq replay` checks past sessions against the policy without installing hooks

  4. 04

    Native hooks for Claude Code, Cursor, Codex, Copilot CLI, Windsurf and Antigravity, plus an OpenClaw plugin and stdio MCP proxy

  5. 05

    Denies curl-pipe-to-shell and secret-carrying outbound calls; asks when a command repeats text the agent read

  6. 06

    Local console on 127.0.0.1 with Check, Live, Agents, Rules and Skills & MCP screens

  7. 07

    `stroq attack` replays 36 known attacks against your policy in a sandbox

02Pricing

Open Source

Free and open-source (Apache-2.0). Runs locally via `npx @stroq/cli` with no account and no telemetry.

03Who it's for

Best for

Developers running Claude Code, Codex or Cursor locally who want to see whether prompt-injected content drove their agent's commands and to guard future sessions without a cloud service.

Not ideal for

Teams needing a true sandbox or centrally managed enterprise policy; matching is text-level and only sees what each agent's hooks expose.

04Verdict
A thoughtful, privacy-first audit and guard for agent sessions with clear documented limits; young and single-maintainer, but shipping weekly.
ai.dosa.dev editorial · Nov 6, 2026
05Recent updates

v0.30.0 (Nov 5, 2026) added the `stroq` session card, detection of tasks slipped into data the agent reads, and a five-screen console; it was posted to Hacker News as a Show HN the same day.

As of

06Pricing & FAQ
Is Stroq free?

Yes - Stroq is Open Source. Free and open-source (Apache-2.0). Runs locally via `npx @stroq/cli` with no account and no telemetry.

Is Stroq open source?

Yes - Stroq is open source. Free and open-source (Apache-2.0). Runs locally via `npx @stroq/cli` with no account and no telemetry.

Who is Stroq best for?

Developers running Claude Code, Codex or Cursor locally who want to see whether prompt-injected content drove their agent's commands and to guard future sessions without a cloud service.

Who is Stroq not ideal for?

Teams needing a true sandbox or centrally managed enterprise policy; matching is text-level and only sees what each agent's hooks expose.

What are the best Stroq alternatives?

The closest Stroq alternatives on ai.dosa.dev are CodeRabbit, Qodo, Snyk Code - all listed under AI Code Review & Security.

Who makes Stroq?

Stroq is developed by Agybay Gibatollayev. It is listed in the AI Code Review & Security category on ai.dosa.dev.

07More in Code Review
08Tags
  • security
  • prompt-injection
  • claude-code
  • codex
  • open-source
Track Stroq in your AI stack

Favorite this tool to revisit it later, or Zap it to contribute to the public vote count.

Compare Code Review Sign in to save
Content on this page is AI-generated. Please verify details with the vendor's website for accuracy. Are you the maker? Get your Featured badge ← Back to directory