skills/delorenj/00_ceiling/hindsight/Gen Agent Trust Hub

hindsight

Pass

Audited by Gen Agent Trust Hub on Jun 23, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Downloads official CLI tools and integrations from vendor-controlled domains and package registries.
  • [REMOTE_CODE_EXECUTION]: Provides instructions for installing the Hindsight memory system through piped shell scripts from official vendor endpoints.
  • [COMMAND_EXECUTION]: Uses the hindsight CLI binary to manage memory banks and perform semantic retrieval operations.
  • [DATA_EXFILTRATION]: Accesses local configuration files (~/.openclaw/openclaw.json) to validate memory governance settings; no unauthorized external transmission of sensitive data was observed.
  • [PROMPT_INJECTION]: Implements a persistent memory architecture that ingests external conversational data and interpolates it into agent prompts. Boundary markers like hindsight_memories tags are used in integration plugins to scope the context. Sanitization routines are present to strip tags before re-storage to prevent recursive injection loops.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 23, 2026, 03:18 AM
Security Audit — agent-trust-hub — hindsight