summarize-session

Pass

Audited by Gen Agent Trust Hub on Nov 11, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill consists entirely of instructional markdown with no executable code, network operations, or suspicious system commands. It follows best practices for data handling, such as verifying information against primary sources (raw logs) and respecting data confidentiality.
  • [INDIRECT_PROMPT_INJECTION]: As a summarization tool, the skill naturally ingests session history which may contain untrusted user content. This creates a surface for potential indirect prompt injection where malicious instructions embedded in the conversation could attempt to bias the summary. However, the skill mitigates this through structured output modes and verification requirements.
  • Ingestion points: Conversation records, file contents, git commits, and command outputs (SKILL.md).
  • Boundary markers: The skill does not define technical delimiters for inputs but enforces a rigid structural output format for the summary.
  • Capability inventory: The skill definition does not include tool calls, scripts, or network capabilities.
  • Sanitization: Explicit instructions are provided to omit sensitive information not authorized for external provision and to discard superseded or retracted statements.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 11, 2026, 02:39 AM