skills/cipherstash/stack/stash-edge/Gen Agent Trust Hub

stash-edge

Pass

Audited by Gen Agent Trust Hub on Sep 27, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the official vendor package @cipherstash/stack from the NPM registry. It also provides guidance on using npm specifiers within Deno environments to ensure proper dependency management.
  • [CREDENTIALS_UNSAFE]: The documentation covers the management of sensitive credentials such as CS_CLIENT_KEY and CS_CLIENT_ACCESS_KEY. It correctly identifies these as secrets and provides instructions for using secure methods of delivery, including platform-specific secret stores (e.g., Supabase secrets or Wrangler secrets) and environment variables, while explicitly advising against committing secrets to source control.
  • [INDIRECT_PROMPT_INJECTION]: The skill describes an interface for processing data through encryption and decryption functions, which represents a potential data ingestion surface.
  • Ingestion points: Untrusted data enters the context via the client.encrypt and client.decrypt methods, as well as their bulk variants mentioned in SKILL.md.
  • Boundary markers: The documented code snippets do not explicitly specify the use of delimiters or boundary markers for the data being processed.
  • Capability inventory: The skill environment supports network operations for key resolution via ZeroKMS and includes CLI-based file writing capabilities through stash env --write for environment configuration.
  • Sanitization: No specific input sanitization or validation logic is detailed in the provided instructions for the data payloads.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 27, 2026, 09:19 PM