periodic-rollup

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted content from the Obsidian vault to generate consolidated notes, creating a surface for indirect prompt injection.
  • Ingestion points: Vault content is read from periodic notes and child note descriptions as specified in references/history-rollup.md and references/periodic-rollup.md.
  • Boundary markers: No explicit delimiters or instructions are used to distinguish vault data from system prompts.
  • Capability inventory: The skill utilizes subprocess calls to execute 'rg' and 'obsidian-cli' and performs file-write operations via 'obsidian-cli create'.
  • Sanitization: No explicit sanitization or filtering is applied to the note content before synthesis.
  • [COMMAND_EXECUTION]: The skill executes shell commands using 'rg' and 'obsidian-cli' for searching and managing vault data. User-supplied arguments like project names are interpolated into these commands, which presents a surface for command injection if input is not properly handled by the agent context.
  • [COMMAND_EXECUTION]: A dynamic context injection '!obsidian daily:path' is used in SKILL.md to retrieve path information during skill initialization. This is a functional use of the platform's features for vault-specific workflows.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 05:59 PM
Security Audit — agent-trust-hub — periodic-rollup