outlook-automation

Pass

Audited by Gen Agent Trust Hub on Apr 3, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted data from incoming emails and calendar invites. 1. Ingestion points: The list, get, and search commands ingest external content from Outlook folders as documented in SKILL.md and reference/cli.md. 2. Boundary markers: No explicit delimiters or instructions are used to isolate ingested email content from the agent's instructions. 3. Capability inventory: The agent can send emails, create meetings, and save attachments as described in the workflows in SKILL.md. 4. Sanitization: No sanitization or filtering of ingested content is mentioned in the provided documentation or JSON output schemas.
  • [COMMAND_EXECUTION]: The skill executes a local Python CLI tool via the uv run command at a hardcoded path (C:/Users/GordonMickel/work/outlookctl). This tool interacts with the Windows COM automation interface to control the Outlook application and performs file system operations such as saving attachments.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 3, 2026, 03:37 AM