graphql-n-plus-one-remediation
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses standard local commands to analyze the repository environment.
- Evidence: Executes
grepto locate resolver patterns andgit diffto identify modified files in a PR context. These commands are used for static analysis of the local project files. - [INDIRECT_PROMPT_INJECTION]: The skill ingests repository code and PR diffs to perform its analysis, which represents a surface for indirect prompt injection.
- Ingestion points: Processes source files in the
resolvers/andschema/directories, as well as PR diff outputs. - Boundary markers: No specific delimiters are used to isolate repository content from the agent's instructions.
- Capability inventory: The skill is designed to modify source code to implement performance fixes.
- Sanitization: The skill lacks explicit sanitization for the code snippets it reads and processes.
Audit Metadata