CheckOps
ArgusRed
ArgusRed, by model lab Cosine, reads your code like an attacker, tries each suspected flaw in an isolated sandbox and reports only exploits it could reproduce, with a fix.
Latest public signals
Not refreshed yet. Refresh reads the public site and, when the entry has a GitHub repository, the latest release.
Introduction
ArgusRed reads your code the way an attacker does, then tries the door. Suspected issues are reproduced in an isolated sandbox, and you receive only the exploits it could actually reproduce, each with a fix.
It is built by Cosine, which describes itself as a model lab rather than a wrapper, and runs in the EU. Its Show HN post described post-training a model that pentests instead of refusing.
Because the fee is refunded when nothing exploitable is confirmed, ArgusRed is an easy way to sanity-check a repository before a launch or a funding round. Treat a clean result as one signal among several rather than a guarantee, and keep static analysis and dependency checks in CI alongside it.
Launch and background: ArgusRed was posted to Hacker News as a Show HN on 20 June 2026.
Features
- Code-reading attacker: Finds candidate flaws in source.
- Sandbox reproduction: Confirms exploits safely.
- Proof-only reports: No unproven maybes.
- Fixes included: Remediation with each finding.
- EU hosting: Runs in the EU.
Pros
- Refund guarantee: $100 refunded unless a real exploit is confirmed.
Cons
- Repo access needed: White-box only.
- New product: 2026 launch.
Usage and Links (Example)
- Connect a repository.
- Pay the $100 scan fee.
- Review reproduced exploits and fixes.
Resources:
Pricing
ArgusRed charges $100 after you connect a repository, fully refunded unless it confirms a real exploit.
Recommended for?
- Small teams wanting a cheap, proof-only code security check.
Conclusion
ArgusRed’s pay-only-if-exploitable model is an unusual and confident offer.
Run it on a repo with known past issues to calibrate.
On CheckOps, ArgusRed is filed under the Security testing stage. Related tools listed here include Semgrep, Snyk, CodeQL, Aikido Security, Checkmarx and Bearer CLI. Which one fits best depends mostly on your stack, your team’s size and skills, and whether you prefer to run tools yourself or pay for a managed service. This page reflects the official website, documentation and repository as they stood on 11 November 2026.
Alternatives
Other tools on CheckOps that cover similar ground. Open a page, or compare one with ArgusRed side by side.
SSemgrepSemgrep is a fast static-analysis engine with a free open-source CLI and a paid platform for cross-file rules, supply chain, and secrets.Compare with ArgusRed
SSnykSnyk CLI scans and monitors your projects for security vulnerabilities.Compare with ArgusRed
CCodeQLCodeQL is GitHub's semantic code analysis engine: it builds a queryable database from source code and runs security queries (MIT-licensed standard library) to find vulnerabilities, powering GitHub code scanning.Compare with ArgusRed
AAikido SecurityAikido Security is an all-in-one developer security platform covering SAST, SCA, secrets, IaC, containers, cloud posture, DAST and runtime protection, with a free Developer plan.Compare with ArgusRed
CCheckmarxCheckmarx agentic application security software platform Combines Hybrid scanning, AI-powered agents, and unified risk intelligence across every attack surface.Compare with ArgusRed
BBearer CLIBearer CLI is a SAST tool that scans source code and analyses data flows to find, filter and prioritise security and privacy risks, with a commercial Bearer Pro from Cycode.Compare with ArgusRed