Aiprox Workflows
ReviewAudited by ClawScan on May 1, 2026.
Overview
The skill is coherent for AIProx workflow automation, but users should notice that it uses a spend token, can schedule recurring paid runs, and relies on an external MCP package/service.
This looks purpose-aligned rather than malicious. Before installing, make sure you trust the AIProx service and MCP package, use a limited spend token, confirm costs and schedules, and double-check any email or webhook destinations.
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.
If the token is misused or a workflow runs more often than intended, it can spend the token balance.
The skill requires a spend token that authorizes paid workflow executions. This is expected for the service, but it is still delegated payment authority.
Set `AIPROX_SPEND_TOKEN` in your MCP server config... Deducted from your Lightning spend token balance.
Use a limited-balance spend token, keep it private, and review estimated costs before creating, running, or scheduling workflows.
Recurring workflows may keep executing and charging sats until the schedule is changed or the workflow is deleted.
The skill intentionally supports persistent scheduled automation. This is disclosed and purpose-aligned, but scheduled jobs can continue running after initial setup.
Chain AI agents into named workflows, schedule them to run on a cron... `@hourly`, `@daily`, `@weekly`
Confirm schedules carefully, monitor run history and sats spent, and delete workflows that should no longer run.
Information included in workflow inputs or outputs may be processed by multiple AIProx agents and sent to the configured recipient or webhook.
Workflow outputs can pass between multiple AI agents and be delivered to external email or webhook destinations. This is core to the product, but users should consider what data is being shared.
Chain AI agents into named workflows... get results delivered by email or webhook.
Avoid sending sensitive data unless appropriate, and verify email addresses and webhook URLs before creating or running workflows.
Installing or running the external MCP package trusts code outside this skill artifact.
The setup example depends on an external npm/npx MCP server package, while the provided artifact set contains only SKILL.md and no reviewed package contents.
"command": "npx", "args": ["aiprox-workflows-mcp"]
Verify the npm package name, publisher, and version before configuring it, and prefer pinned versions where available.