cipher-type-planner

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's instructions are focused on guiding an AI agent through a standard software development lifecycle (SDLC) workflow within the Bitwarden monorepo. It utilizes internal design patterns (e.g., using SshKey as a reference) and emphasizes feature-flag gating for safety and backward compatibility. No evidence of prompt injection, data exfiltration, or unauthorized command execution was found.
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for indirect prompt injection as it processes user-provided requirements and codebase content to generate a plan. However, the risk is categorized as safe/low because the skill's output is a static markdown plan intended for human review, and it does not automate the execution of the proposed changes.
  • Ingestion points: User-provided answers to implementation questions and source code files from the local repository (e.g., libs/common/src/vault/enums/cipher-type.ts).
  • Boundary markers: Absent; the instructions do not specify delimiters for user-provided field names or descriptions in the plan output.
  • Capability inventory: The skill explores the local filesystem and writes a text-based plan file; it does not possess network exfiltration or shell execution capabilities beyond standard file access.
  • Sanitization: None specified; the skill relies on the underlying model's inherent safety filters and developer review of the generated plan.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 03:54 PM
Security Audit — agent-trust-hub — cipher-type-planner