Trusted — Risk Score 0/100
Last scan:1 day ago Rescan
0 /100
openclaw-memory
Lightweight file-based memory system for single-user AI agents. No databases, no APIs — just markdown files with tiered loading, grep tagging, and automatic archival.
Clean file-based memory system with straightforward shell scripts performing only documented file operations. No network access, credential harvesting, obfuscation, or hidden functionality detected.
Skill Nameopenclaw-memory
Duration25.7s
Enginepi
Safe to install
No action required. This is a legitimate file-based memory management skill.
ResourceDeclaredInferredStatusEvidence
Filesystem READ,WRITE READ,WRITE ✓ Aligned SKILL.md declares tiered file loading; scripts implement file reads/writes
Shell WRITE WRITE ✓ Aligned Scripts/atomic-write.sh, scripts/archive-old-episodic.sh, scripts/health-check.s…
Network NONE NONE No network calls in any script
Environment NONE NONE Scripts use only hardcoded paths; no env iteration
credential NONE NONE No credential access, SKILL.md explicitly prohibits credentials in episodic note…

File Tree

5 files · 13.5 KB · 399 lines
Markdown 2f · 267L Shell 3f · 132L
├─ 📁 references
│ └─ 📝 design-rationale.md Markdown 61L · 2.5 KB
├─ 📁 scripts
│ ├─ 🔧 archive-old-episodic.sh Shell 23L · 806 B
│ ├─ 🔧 atomic-write.sh Shell 16L · 377 B
│ └─ 🔧 health-check.sh Shell 93L · 2.4 KB
└─ 📝 SKILL.md Markdown 206L · 7.5 KB

Security Positives

✓ SKILL.md explicitly prohibits credentials in episodic notes (Safety Rules)
✓ Scripts use safe patterns: atomic writes (tmp→sync→mv), date-based archival, invariant checks
✓ No network access in any script — purely local file operations
✓ No obfuscation, base64 encoding, or eval patterns
✓ No credential harvesting or environment variable enumeration
✓ Hardcoded paths in scripts are consistent with the documented memory directory structure
✓ Simple, auditable bash scripts with no external dependencies