Detect unsafe GraphQL operation document construction with user input tracing.
npx skills add https://github.com/utkusen/sast-skills --skill sast-graphql
Please help me install this Agent Skill. Skill: sast-graphql Source: https://github.com/utkusen/sast-skills/tree/main/sast-files/.agents/skills/sast-graphql Command: npx skills add https://github.com/utkusen/sast-skills --skill sast-graphql
GraphQL injection vulnerabilities occur when user-controlled data is embedded into GraphQL operation strings rather than passed solely via variables. This skill provides a structured three-phase workflow to recon GraphQL usage, verify potential injection sites in batches, and merge results into a single report, enabling secure GraphQL implementations.
Run the GraphQL injection detection workflow by first performing the sast-analysis to generate architecture context, then execute sast-graphql to produce the phased findings.
High-intent search queries and answers about installing and using this skill.
To detect GraphQL injection vulnerabilities, this skill identifies unsafe construction of GraphQL operation documents and traces user input to those sites. It applies a three-phase workflow to locate assembly points, verify taint flow in batches, and consolidate findings into a remediation report.
GraphQL injection occurs when user-controlled data is embedded directly into operation strings rather than passed safely via variables. This skill identifies these unsafe document construction patterns by tracing user input to the sites where dynamic operation text is built.
You run static analysis for unsafe GraphQL document construction by executing a three-phase workflow: recon to locate assembly sites, batched verification to trace taint from user input, and a merge phase to produce a consolidated findings report for remediation.
Yes, you need an existing sast/architecture.md file to scan for GraphQL injection. This skill is architecture-aware and requires that context to accurately target and prioritize the detection of unsafe operation document assembly across your codebase.
The GraphQL injection detection workflow outputs its consolidated findings to sast/graphql-results.md. This file provides the verified results of the three-phase analysis for remediation planning and security auditing.
Yes, you can trace user input to GraphQL operation documents in parallel. The verification phase of the workflow processes potential injection sites in parallel batches to validate taint flow efficiently before merging the results.
utkusenIdentify and document XML parsing points vulnerable to XXE across a codebase.
utkusenMap user-controlled data to dangerous sinks and write RCE findings to rce-results.md.
iiammaeMap SARIF and weAudit findings onto Trailmark code graphs for security audits.
OneWave-AIReview code for security vulnerabilities, performance bottlenecks, and maintainability issues.
