Cron Cost Guard

PassAudited by ClawScan on May 10, 2026.

Overview

This is a coherent, instruction-only cron cost audit checklist with no code or credentials, but it does ask users to inspect local logs/context files and may remove a targeted cron job during remediation.

This skill appears safe to use as an audit checklist. Before following remediation steps, verify cron job IDs, review local log output for sensitive information, and be deliberate when moving persistent context files because those changes can affect future agent behavior.

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.

What this means

Following the remediation can stop a runaway cost loop, but it can also disable a scheduled task if the wrong job ID is used.

Why it was flagged

This is an explicit mutation of persistent scheduled automation. It is purpose-aligned and scoped to an identified runaway cron job, but users should confirm the target before removal.

Skill content
Remove the offending cron job immediately:

```
cron remove (jobId: "<id>")
```
Recommendation

Confirm the cron job ID, owner agent, and expected business impact before removing or recreating the job.

What this means

The agent may inspect recent local gateway log lines, which can reveal operational details such as session IDs, errors, or model-switch behavior.

Why it was flagged

The skill documents a local shell command that reads OpenClaw logs. This is relevant to diagnosing retry loops and is not hidden, but users should expect local command and log access.

Skill content
```bash
tail -200 ~/.openclaw/logs/gateway.log | grep -i "model switch\|retry\|error\|LiveSession"
```
Recommendation

Run the log command only in the intended OpenClaw environment and review any sensitive log output before sharing it elsewhere.

What this means

Moving large context files may reduce token usage, but it can also alter the agent's future behavior if important instructions are no longer injected by default.

Why it was flagged

The skill audits and recommends reorganizing persistent or injected agent context files. This is purpose-aligned for reducing token cost, but it can change what future agent sessions receive as context.

Skill content
wc -c MEMORY.md SOUL.md AGENTS.md TOOLS.md USER.md QUEUE.md ... Move large files ... to `references/` for on-demand reading.
Recommendation

Review context files before moving them, keep critical safety or operating instructions in the right place, and document any changes to persistent agent context.