continuous-monitoring

name	continuous-monitoring
description	Set up automated recurring code audits and optimization monitoring with cron jobs
version	1.0.0
author	kai-agent
metadata	{"kai":{"tags":["kai","monitoring","automation","cron","scheduled"]}}

Continuous Monitoring

Set up automated, recurring security scans and evolution monitoring. The agent runs on a schedule, checks for new results, and reports findings to the team.

Use cases

1. Scheduled code audits: Audit repos weekly or on each branch update
2. Optimization progress tracking: Monitor long-running optimizations and report milestones
3. Multi-repo dashboard: Regular status reports across all repos in a workspace

Setting up scheduled scans

Weekly security scan

schedule_cronjob:
  name: "weekly-security-scan-[repo-name]"
  schedule: "0 9 * * 1"  # Every Monday at 9 AM
  task: |
    Run a security scan on [repo] in workspace [workspace].
    Use tier [tier-id].
    When complete, post a summary of findings to Slack.
    Compare with last week's scan and highlight new vulnerabilities.

Daily evolution check

schedule_cronjob:
  name: "evolution-monitor-[evolution-id]"
  schedule: "0 */4 * * *"  # Every 4 hours
  task: |
    Check progress of evolution [evolution-id].
    If there's a significant fitness improvement since last check, notify the team.
    If the evolution has completed, fetch the best program and post a summary.

Monitoring patterns

Scan comparison over time

For each scheduled scan, save results to memory:

memory: add
key: scan-results-[repo]-[date]
value: { total: N, critical: N, high: N, medium: N, low: N, new: [...], resolved: [...] }

Then compare with previous scan to generate a delta report:

• New vulnerabilities introduced
• Vulnerabilities resolved
• Severity trend (improving/degrading)

Evolution convergence tracking

Track fitness scores across iterations:

• Plot fitness over time (describe trend)
• Detect stagnation (fitness flat for 20+ iterations)
• Suggest abort if convergence is reached early

Tips

• Use descriptive cronjob names that include the repo/workspace
• Set reasonable schedules (don't audit every hour, it's wasteful)
• Always include a notification action (Slack message, log update)
• Save monitoring history in memory for trend analysis
• Use list_cronjobs and remove_cronjob to manage active monitors