session-index

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS] (MEDIUM): The skill requires the installation of the claude-session-index package via pip. This package and its author ('Lee Fuhr') are not among the trusted sources or organizations specified in the security framework.
  • [COMMAND_EXECUTION] (LOW): The skill instructions direct the agent to construct and execute shell commands (e.g., sessions "keywords") by interpolating user input directly into command strings, which could lead to command argument injection if keywords are not properly sanitized.
  • [Indirect Prompt Injection] (LOW): The skill processes untrusted data from previous Claude Code sessions stored in ~/.session-index/sessions.db. Malicious instructions embedded in past conversations could be re-activated during the 'synthesis' phase.
  • Ingestion points: Session data is read from ~/.session-index/sessions.db and ~/.claude/session-topics/.
  • Boundary markers: No specific delimiters or 'ignore' instructions are provided when interpolating session history into the synthesis task.
  • Capability inventory: Includes bash command execution and the ability to spawn sub-tasks using the haiku model.
  • Sanitization: No sanitization or validation of the retrieved session content is mentioned.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 06:15 PM