AI Workflow Automation Expert
PassAudited by ClawScan on May 1, 2026.
Overview
This is an instruction-only automation guide; it does not install code or request credentials, but users should be mindful that its examples involve scheduled, multi-agent, and persistent workflows.
This skill appears safe as an instruction-only automation guide. Before following its templates, review any added skills such as email, API gateway, cron, social posting, autonomous tasks, or shared memory, and keep high-impact workflows manual or approval-gated until tested.
Findings (3)
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 user implements these examples, an agent could run recurring tasks on a schedule.
The reference shows how to create scheduled agent turns. This is central to workflow automation, but scheduled agent activity can continue without immediate user presence.
"schedule": { "kind": "cron", "expr": "0 9 * * *" }, "payload": { "kind": "agentTurn", "message": "Check inbox and summarize" }Use narrow schedules, isolated sessions where appropriate, logging, and manual review for sensitive or high-impact tasks.
Poorly scoped agent-to-agent messages could send task data to the wrong worker or reuse data in unintended contexts.
The skill documents message-based coordination between agents. This is expected for multi-agent workflows, but users should define identity, scope, and data boundaries when implementing it.
Use sessions_send for inter-agent messaging... target_session: "worker-agent"
When building multi-agent workflows, explicitly define which agents may receive which data and avoid sending secrets or private content unless necessary.
If implemented carelessly, workflow memory could preserve private data or outdated instructions across later tasks.
The reference recommends persistent state and memory for multi-agent workflows. This is purpose-aligned, but persistent memory can retain sensitive or stale instructions if not managed.
Use `memory/` directory for persistent state; Use `MEMORY.md` for long-term knowledge
Keep persistent memory scoped to the workflow, avoid storing secrets, and periodically review or clear saved state.