Documentation contains a destructive delete command without an explicit confirmation gate.
Warn
- Code
- suspicious.destructive_delete_command
- Location
- README.md:88
- Evidence
rm -rf ~/.openclaw/memory-braid
Security audit
memory-braid
Security checks across malware telemetry and agentic risk
Overview
This appears to be a real memory plugin that captures and recalls user memories, with no obvious unrelated permissions, but users should understand it can store conversation-derived memories and may use external AI services if configured.
Install this only if you want a plugin that can remember selected information across sessions. Review its configuration before enabling cloud/ML features, because captured memory text may be stored in Mem0 and optional OpenAI-based extraction can send text to OpenAI when configured. The main transparency caveat is that the registry declares no environment variables even though optional external-provider features may use API keys.
VirusTotal
VirusTotal engine telemetry is currently stale for this artifact.
Static analysis
Detected: suspicious.destructive_delete_command, suspicious.env_credential_access, suspicious.exposed_secret_literal
Environment variable access combined with network send.
Critical
- Code
- suspicious.env_credential_access
- Location
- src/entities.ts:45
- Evidence
process.env.OPENCLAW_STATE_DIR?.trim() ||
Environment variable access combined with network send.
Critical
- Code
- suspicious.env_credential_access
- Location
- src/extract.ts:253
- Evidence
const key = process.env.OPENAI_API_KEY;
File appears to expose a hardcoded API secret or token.
Critical
- Code
- suspicious.exposed_secret_literal
- Location
- src/mem0-client.ts:532
- Evidence
const apiKey = [REDACTED]?.trim() || process.env.MEM0_API_KEY;