ClawHub
Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Shared Memory OS

v1.0.5

Shared memory governance for multi-agent OpenClaw workspaces — with tiered memory, heartbeat maintenance, review cycles, conflict handling, and controlled ev...

⭐ 0· 82·0 current·0 all-time
byQihong@zqh2333
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Benign
medium confidence
✓
Purpose & Capability
The name and description promise shared-memory governance; the SKILL.md only requires agents to read and update workspace memory files and governance docs. No unrelated binaries, credentials, or external services are requested.
ℹ
Instruction Scope
The instructions direct agents to read many workspace files (SOUL.md, USER.md, MEMORY.md, memory/*, HEARTBEAT.md, etc.) which is appropriate for a shared-memory governance layer. This is broad by design — ensure agents are only allowed to access private files (e.g., memory/private-secrets.md) when explicitly required.
✓
Install Mechanism
No install spec and no code files — instruction-only skill. Nothing is written to disk or executed as part of an installer.
ℹ
Credentials
The skill requests no environment variables or credentials. However the included references discuss a memory/private-secrets.md template for storing secrets in workspace files — storing secrets in plaintext files is risky. The skill itself instructs to read private-secrets.md only on explicit credentialed tasks, which is a mitigating control.
✓
Persistence & Privilege
always:false (default). The skill does not request permanent platform-wide privileges or to modify other skills' configs. It can be invoked by agents per normal platform behavior.
Assessment
This skill is coherent with its stated purpose and is instruction-only, but review your workspace before installing: 1) Check whether you already have a memory/private-secrets.md or other files that contain unencrypted secrets — if so, move secrets to a secure secret store and avoid letting agents read them by default. 2) Decide which agent sessions are allowed to access private memory (limit to credentialed/private sessions). 3) Because the skill's source/homepage are absent, verify the skill content and any local modifications before publishing or enabling it in production. 4) Test in a staging workspace to confirm agents follow the 'do not read private-secrets.md by default' rule and that heartbeat/auto-promotion behavior is acceptable for your workflows.

Like a lobster shell, security has layers — review code before you run it.

latestvk9782t5g652c6dapgqk6badrc1842sqe

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments