memorb-forgetter

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface due to the way it ingests and processes external data.
  • Ingestion points: File targets and category metadata are derived from external memorb-lint reports or user-supplied archive requests in SKILL.md.
  • Boundary markers: The instructions lack explicit delimiters or markers to isolate these external inputs from the agent's logic.
  • Capability inventory: The skill has high-impact capabilities including physical file relocation and vault-wide text search-and-replace for link integrity.
  • Sanitization: There is no mention of escaping or sanitizing the input variables to prevent path traversal or unintended file access outside the memorbs/ directory.
  • [COMMAND_EXECUTION]: The skill executes file system modifications using logic equivalent to shell commands (mv). Although the operations are described within a specific vault structure, performing these actions based on externally-influenced variables without rigorous path validation represents a minor security risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 09:48 AM
Security Audit — agent-trust-hub — memorb-forgetter