gs-bun-aws-lambda

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches the official Bun runtime installer from 'https://bun.sh/install' and binaries from the 'oven-sh' GitHub organization. These are official and well-known sources.
  • [REMOTE_CODE_EXECUTION]: Executes the Bun installer by piping a remote script to bash within the Dockerfile and deployment references. This is the official and recommended installation method for the Bun project.
  • [COMMAND_EXECUTION]: Employs shell scripts to implement the AWS Lambda Runtime API loop in custom runtimes, using 'curl' to manage invocations and responses.
  • [PROMPT_INJECTION]: The skill provides templates that ingest untrusted data from Lambda event sources, creating a surface for indirect prompt injection.
  • Ingestion points: 'src/functions/create-category.ts' (event.body) and 'src/functions/process-queue.ts' (record.body).
  • Boundary markers: Absent.
  • Capability inventory: Shell calls in bootstrap scripts, network operations via 'fetch', and file access via 'Bun.write'.
  • Sanitization: Not present in the code snippets.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 02:17 AM
Security Audit — agent-trust-hub — gs-bun-aws-lambda