compress-session

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes historical session logs and current conversation context to generate reports, which presents a surface for indirect prompt injection where adversarial content in logs could influence the agent during distillation.
  • Ingestion points: The skill instructions specifically direct the agent to read logs from quality_reports/session_logs/ and the current conversation content.
  • Boundary markers: While the skill provides a structured markdown template for its output, it does not define strict delimiters or instruct the agent to ignore potentially malicious commands within the ingested text.
  • Capability inventory: The skill is granted access to the Bash, Write, Read, Glob, and Grep tools.
  • Sanitization: The skill contains no instructions for filtering, sanitizing, or escaping the content retrieved from session logs before inclusion in the final report.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 03:57 PM
Security Audit — agent-trust-hub — compress-session