skills/shipshitdev/skills/session-end/Gen Agent Trust Hub

session-end

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill orchestrates a local session-management workflow without any network operations, remote script downloads, or data exfiltration. All data storage is directed to a hidden local directory.
  • [NO_CODE]: The skill contains no executable scripts or binary code; its functionality relies entirely on natural language instructions and configuration metadata.
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by ingesting and potentially reloading session context that may include untrusted data from external files or user tasks.
  • Ingestion points: Session context capturing file summaries and task logs (SKILL.md).
  • Boundary markers: Absent; there are no specific delimiters defined to segregate recorded session data from agent system instructions.
  • Capability inventory: No direct subprocess, network, or file-write capabilities are defined within this orchestration skill.
  • Sanitization: Absent; the skill does not perform validation or sanitization on recorded session data.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 09:46 AM