Quality

Fix Schema updates Zod schemas to support new session log data formats while maintaining backward compatibility. Use it when session log JSON parsing fails and a pull request is needed.

Detects business logic vulnerabilities in codebases using a three-phase approach combining threat modeling, parallel verification of attack scenarios, and result merging. Use when finding logic flaws, workflow bypasses, or abuse-of-function bugs.

Detect Cross-Site Scripting vulnerabilities in codebases via a three-phase SAST approach: recon, batched verification with subagents, and merge. Use when analyzing web application source code for XSS bugs.

Detect GraphQL injection vulnerabilities by tracing user input to unsafe operation document assembly sites. Use for security audits and code reviews of GraphQL APIs.

Detects hardcoded sensitive data (API keys, tokens, passwords) in publicly accessible code like frontend JavaScript, mobile apps, and HTML templates. Uses recon, verify, and merge phases. Requires architecture analysis first.

Detect IDOR vulnerabilities by identifying endpoints lacking ownership checks, using automated recon, parallel batch verification, and merge. Ideal for code security audits targeting authorization bypass.

Detect insecure file upload vulnerabilities in a codebase using a three-phase approach: discovery, batched verify, and merge. Use when asked to find file upload, unrestricted upload, or extension bypass bugs.

Detects insecure JWT implementations in codebases. Uses a two-phase approach: mapping JWT issuance/verification sites, then checking for weaknesses such as algorithm confusion, missing signature verification, weak secrets, and missing claim validation. Ideal for finding token forgery or authentication bypass bugs.

Analyzes codebases for insecure JWT implementations: maps token lifecycle, then checks verification sites for algorithm confusion, missing signature verification, weak secrets, header injection, and missing claim validation.

Detects missing authentication and broken function-level authorization vulnerabilities via a three-phase recon-batch-merge approach. Use when asked to find unauthenticated endpoints or vertical privilege escalation bugs.

Detects path traversal vulnerabilities in a codebase using a three-phase approach: recon, batched verify, and merge. Use when auditing code for file disclosure or directory traversal bugs.

Detects remote code execution vulnerabilities (command injection, eval injection, unsafe deserialization) by finding dangerous sinks and tracing user input with parallel subagents. Use when auditing code for critical injection flaws.