OpenClaw Ops
AdvisoryAudited by Static analysis on Apr 30, 2026.
Overview
No suspicious patterns detected.
Findings (0)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
A trusted agent using this skill can affect whether the OpenClaw Gateway runs correctly.
The skill instructs an agent to run local shell and service-management commands that can change or restart the gateway. This is disclosed and central to the rescue/ops purpose.
Shell access on the same machine as the OpenClaw Gateway ... systemctl --user restart openclaw-gateway
Use it only with a trusted rescue agent, review proposed shell commands, and require explicit confirmation for restarts, upgrades, deletes, or configuration edits.
If used by an untrusted agent or in an unsafe session, OpenClaw tokens or account access could be exposed or misused.
The skill expects access to local token-bearing environment/config files. This is purpose-aligned for diagnosing auth failures, but it is sensitive credential access.
The agent will access env files containing tokens (but is instructed never to print them)
Install only for agents you trust with OpenClaw credentials, avoid displaying token values, and rotate tokens if they are accidentally printed or shared.
The rescue agent may inspect or remove local session history while performing maintenance.
OpenClaw sessions and transcripts may contain private conversation history or agent state. The skill discloses this access and says deletion requires confirmation.
Read/write access to `~/.openclaw/` (config, agents, sessions) ... Delete old session transcripts (only with user confirmation)
Confirm exactly which sessions or transcripts will be read or deleted, and keep backups before cleanup operations.
Installing directly from the repository could fetch content that differs from the reviewed ClawHub artifact.
The README describes a user-directed install from an external repository rather than a pinned artifact. No hidden scripts are present in the supplied artifacts.
Install the openclaw-ops skill from https://github.com/dinstein/openclaw-ops-skill ... The agent will download `SKILL.md`
Prefer the reviewed registry package or verify the exact Git commit/content before asking an agent to download and install it.