graphql-idor-via-introspection-leak
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely informational and instructional. It provides guidance on identifying Insecure Direct Object Reference (IDOR) vulnerabilities in GraphQL APIs through introspection leaks.
- [NO_CODE]: The skill consists of markdown documentation with GraphQL query examples and simple shell commands for demonstration (e.g., base64 encoding). There are no scripts, automated execution patterns, or external dependencies introduced by this skill.
- [REMOTE_CODE_EXECUTION]: No remote code execution patterns or instructions to download and execute external scripts were found.
- [DATA_EXFILTRATION]: There are no commands that access sensitive local files or attempt to exfiltrate data to external servers.
- [OBFUSCATION]: Technical examples of Base64 encoding are provided for educational purposes to explain Relay Global IDs, but no malicious obfuscation or hidden content is present.
- [PROMPT_INJECTION]: The skill does not contain instructions that attempt to override agent behavior or bypass safety guardrails.
Audit Metadata