canvas-present

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local Python scripts, specifically scripts/canvas_template.py and scripts/canvas_validate.py, to generate and verify Obsidian Canvas structures based on provided parameters.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection when utilizing the /canvas present from [notes] operation, which parses untrusted external content into the agent's context.
  • Ingestion points: Markdown notes and wiki pages read by the agent to extract slide content (H1/H2 headings, paragraphs, and lists).
  • Boundary markers: No delimiters or instructions are provided to the agent to ignore potentially malicious instructions embedded within the source notes.
  • Capability inventory: The skill has the ability to execute local subprocesses (Python scripts) and perform file system modifications using the Edit tool.
  • Sanitization: There is no evidence of sanitization or validation of the extracted markdown content before it is processed and written to the canvas file.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 05:07 PM