obsidian

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill accesses system environment variables and configuration files to determine the vault path.
  • The instructions direct the agent to check the OBSIDIAN_VAULT_PATH environment variable and the ~/.hermes/.env file.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection attack surface by processing untrusted data with file-system write capabilities.
  • Ingestion points: The agent reads note content using read_file and searches for content using search_files (found in SKILL.md).
  • Boundary markers: There are no instructions provided to the agent to treat note content as untrusted or to ignore embedded instructions.
  • Capability inventory: The skill utilizes write_file, patch, and terminal tools, allowing for file modification and command execution across the filesystem.
  • Sanitization: There is no evidence of sanitization or validation of the data retrieved from notes before it is used in subsequent operations.
  • [COMMAND_EXECUTION]: The skill utilizes the terminal for path resolution and system utilities.
  • It uses the terminal tool to verify the existence of the vault path and resolve environment variables.
  • It executes shell commands for date arithmetic (e.g., date -v+1d) to manage daily note structures.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 07:56 AM
Security Audit — agent-trust-hub — obsidian