spark-recipe-end-of-day

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources which could contain malicious instructions.
  • Ingestion points: The output of spark emails (senders, subjects) and spark events (meeting titles) in SKILL.md is processed by the agent to generate a summary.
  • Boundary markers: None. The skill does not use specific delimiters or instructions to ignore embedded commands in the email/event data.
  • Capability inventory: The skill uses spark CLI commands to read data but does not perform file writes, network exfiltration, or dynamic code evaluation.
  • Sanitization: There is no explicit sanitization or filtering of the text content retrieved from the Spark CLI before it is presented to the agent for summarization.
  • [COMMAND_EXECUTION]: The skill executes shell commands using the spark CLI utility to fetch email and calendar information. These commands use static filters and do not interpolate unvalidated user input, minimizing command injection risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 07:08 PM
Security Audit — agent-trust-hub — spark-recipe-end-of-day