taxue-save

Pass

Audited by Gen Agent Trust Hub on Jun 27, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill implements a state restoration mechanism that creates an indirect prompt injection surface by re-injecting historical data into the agent's context.
  • Ingestion points: Historical session data is retrieved from ~/.taxue/sessions/ and behavioral patterns are read from ~/.taxue/growth/patterns.md (as described in SKILL.md).
  • Boundary markers: The data is structured using YAML frontmatter and Markdown headers, but the instructions do not include specific delimiters or directives to ignore potentially malicious instructions embedded in the saved state.
  • Capability inventory: The skill performs filesystem operations including directory creation and read/write access to files within the user's home directory (~/.taxue).
  • Sanitization: No explicit sanitization, validation, or escaping of the retrieved session content is defined before it is interpolated into the agent's prompt.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 27, 2026, 11:25 AM
Security Audit — agent-trust-hub — taxue-save