bear-notes

Warn

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill metadata specifies the installation of a CLI tool via go install github.com/tylerwince/grizzly/cmd/grizzly@latest. This repository is owned by an unverified third-party user, introducing a risk of supply chain compromise or unvetted code updates.
  • [COMMAND_EXECUTION]: The skill performs shell command execution to interact with the Bear application, utilizing the grizzly binary with various flags and piped inputs.
  • [CREDENTIALS_UNSAFE]: The skill interacts with and instructs users to store a Bear API token in a plaintext file at ~/.config/grizzly/token. The use of sensitive file paths for credential storage increases the risk of local data exposure.
  • [PROMPT_INJECTION]: The skill ingests untrusted data from Bear notes via grizzly open-note and grizzly tags (as documented in SKILL.md) without boundary markers or sanitization. This creates a surface for indirect prompt injection where malicious content in a note could influence agent behavior. The skill includes write capabilities through note creation and text appending commands.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 6, 2026, 04:19 PM