cass-memory

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill operates by executing a local command-line interface tool named cm. It provides instructions for various subcommands including initialization (cm init), context retrieval (cm context), pattern extraction (cm reflect), and system health checks (cm doctor).
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and replay data from past sessions, which is then used to influence the agent's behavior in current tasks.
  • Ingestion points: The agent ingests potentially untrusted data from the output of cm context (rules, anti-patterns, history snippets), cm reflect (extracted patterns), and cm playbook get (rule details) as defined in SKILL.md.
  • Boundary markers: The instructions do not define delimiters or specific warnings to ignore instructions that might be embedded within the retrieved context or history snippets.
  • Capability inventory: The agent is authorized to execute shell commands via the cm CLI and interpret its output to inform task approaches.
  • Sanitization: There is no mention of sanitization, validation, or filtering of the historical data or user-contributed rules before they are re-interpolated into the agent's prompt context.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 08:25 PM
Security Audit — agent-trust-hub — cass-memory