Releases·Tool Release
Show HN: Security Cards – Reducing insecure AI-generated code by 72%
Reware Labs released Security Cards, an open-source set of targeted security guidance covering 80+ widely used libraries across 13 programming languages. Internal evaluation reports the cards reduce insecure code generation by up to 72.3% when used with Claude Code running Opus 4.7, addressing agents that produce functionally correct but insecure code.
6 engagement1 sourceThu, Sep 10, 2026, 12:03 PM
AI Overview
- Reware Labs open-sourced Security Cards offering security guidance for 80+ widely used libraries across 13 programming languages.
- Security Cards cut insecure code generation by up to 72.3% in Claude Code with Opus 4.7, per the project's evaluation.
- The cards are available on GitHub and securitycards.rewarelabs.com, with Reware Labs soliciting feedback on which libraries to support next.
Related
Products·launched
AI Code Security Cards: guide coding agents to use libraries securely
AI Code Security Cards is an open-source security knowledge layer that provides coding agents with library- and version-specific guidance to generate safer code. It addresses the problem of LLM-generated code lacking security awareness by supplying contextual security knowledge. Aimed at developers and security teams using AI coding agents.
AI Code Security Cards is an open-source security knowledge layer that provides coding agents with library- and version-specific guidance to generate safer code. It addresses the problem of LLM-generated code lacking security awareness by supplying contextual security knowledge. Aimed at developers and security teams using AI coding agents.
1 engagement1 sourcehackernewsTue, Aug 4, 2026, 02:12 PM
Products·launched
Open-source tool to fix security issues in AI-generated code
This open-source tool helps developers identify and fix security vulnerabilities in AI-generated code, addressing the high rate of security issues (40%+) found in such code. It uses LLMs to analyze and suggest fixes, targeting developers and security teams.
This open-source tool helps developers identify and fix security vulnerabilities in AI-generated code, addressing the high rate of security issues (40%+) found in such code. It uses LLMs to analyze and suggest fixes, targeting developers and security teams.
0 engagement1 sourceredditWed, Aug 5, 2026, 09:35 AM
Products·launched
CodeInspectus: open-source security scanner for AI-generated code
CodeInspectus is a fully open-source, local security scanner that checks AI-generated code for vulnerabilities. It covers 32 checks (13 AI-specific + 19 SAST) and 200+ secret/API-key patterns, catching issues like hardcoded secrets in client-side code, exposed API keys, and insecure RLS policies. It helps developers secure projects built with LLM-generated code.
CodeInspectus is a fully open-source, local security scanner that checks AI-generated code for vulnerabilities. It covers 32 checks (13 AI-specific + 19 SAST) and 200+ secret/API-key patterns, catching issues like hardcoded secrets in client-side code, exposed API keys, and insecure RLS policies. It helps developers secure projects built with LLM-generated code.
1 engagement1 sourceredditSun, Jul 12, 2026, 03:22 PM
Developments·Paper
Paper proposes just-in-time pipeline to detect and fix vulnerabilities in AI-generated code
A new arXiv paper describes an automated pipeline that generates Python code from LLMSecEval prompts, scans it with CodeQL and Bandit alongside an independent Code Validator LLM, and enriches findings with MITRE ATT&CK techniques and CWE examples. The approach aims to detect, fix, and verify security issues at development velocity, grounding remediation in real-world threat context.
A new arXiv paper describes an automated pipeline that generates Python code from LLMSecEval prompts, scans it with CodeQL and Bandit alongside an independent Code Validator LLM, and enriches findings with MITRE ATT&CK techniques and CWE examples. The approach aims to detect, fix, and verify security issues at development velocity, grounding remediation in real-world threat context.
0 engagement1 sourcearxivMon, Aug 17, 2026, 07:09 AM
Products·launched
vibeArchitecture: AI code generator that enforces OWASP security rules
vibeArchitecture is a tool that prevents AI-generated code from shipping common security vulnerabilities (OWASP Top 10) by enforcing security rules during code generation. It solves the problem of LLMs producing insecure code by default, targeting developers using AI coding assistants.
vibeArchitecture is a tool that prevents AI-generated code from shipping common security vulnerabilities (OWASP Top 10) by enforcing security rules during code generation. It solves the problem of LLMs producing insecure code by default, targeting developers using AI coding assistants.
1 engagement1 sourceredditThu, Jul 16, 2026, 04:37 PM