Greg Eisenberg
Analysis
This is an instruction-only brainstorming/style skill with no code or requested credentials, though its reference notes discuss high-trust agent setup practices users should apply carefully.
Findings (4)
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.
Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.
**Dangerously skip permissions mode** — bypasses confirmation prompts (use with caution)
The reference material mentions a mode that reduces user confirmation prompts. It is clearly labeled with caution and is part of Claude Code workflow notes, not an instruction for this skill to enable it automatically.
**Cron health check** — monitors if cron jobs failed and re-triggers them ... Keep heartbeat instructions lean — it runs constantly
The reference describes long-running heartbeat/cron automation. This is relevant to OpenClaw automation guidance and not implemented by the skill, but it is persistent behavior users should scope carefully.
Checks whether tool use, credentials, dependencies, identity, account access, or inter-agent boundaries are broader than the stated purpose.
**Chrome Relay** — Chrome extension on your main browser, lets OpenClaw take over your existing logged-in sessions. Less recommended
The reference discusses browser/session delegation, which is high-trust account access. It also says this method is less recommended and later advises dedicated agent-owned accounts, so this is a cautionary note rather than hidden credential use.
Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.
`memory.md` doesn't exist by default — you need to tell OpenClaw to create it ... Add memory autosave to heartbeat — every 30 minutes, save current session to memory
The reference recommends persistent memory and periodic autosave, which can retain user context across sessions. The skill itself does not implement storage, but this is sensitive setup advice users should understand.