graincrawl
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for building, testing, and managing a Go application using
go build,go test,go mod, andmakecommands. These are standard operations for the skill's stated purpose of project maintenance. - [EXTERNAL_DOWNLOADS]: The skill references a Homebrew tap (
vincentkoc/tap/graincrawl) for installation and testing. The repository belongs to the skill's author, representing an expected vendor resource. - [INDIRECT_PROMPT_INJECTION]: The workflow involves reading documentation files from the local project repository which could serve as an attack surface for indirect prompt injection if the repository contains untrusted content.
- Ingestion points: The agent is instructed to read
AGENTS.md,README.md,SPEC.md, anddocs/security.mdfrom the graincrawl checkout inSKILL.mdStep 1. - Boundary markers: Absent.
- Capability inventory: Shell command execution (
go,git,brew), file system access, and binary compilation. - Sanitization: Absent; the skill does not specify validation or sanitization for the content of the documentation files.
Audit Metadata