weekly-summary

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes a local command-line utility named sessions to process session history and manage memory candidates. It specifically calls sessions memory mine and sessions memory pending to automate the extraction of relevant information for the weekly summary.
  • [INDIRECT_PROMPT_INJECTION]: The skill retrieves and summarizes historical session data which includes past user and agent messages. This creates a surface where instructions embedded in previous coding sessions could potentially influence the current summarization task.
  • Ingestion points: Historical session content is ingested via the get_activity_digest and get_session_messages tools.
  • Boundary markers: The skill does not implement specific delimiters or 'ignore' instructions when processing historical message content.
  • Capability inventory: The skill possesses the capability to execute local CLI commands and read session history across all projects.
  • Sanitization: Historical session data is processed directly without explicit sanitization or filtering of potential instructional content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 07:32 AM