Every weekday morning at 8:00 UTC, investigate Datadog logs and metrics and post a SHORT report to Slack channel #C0AGCTEREE9 (channel ID C0AGCTEREE9).

Use the available secrets/environment only; never print or expose secret values:

• Datadog: DATADOG_API_KEY, DD_APP_KEY, DD_SITE

• Slack: SLACK_BOT_TOKEN

Reporting window:

• Tuesday-Friday runs: analyze the last 24 hours.

• Monday runs: analyze the last 72 hours so the weekend is included.

Investigation requirements:

1. Query Datadog logs and metrics/APM/monitors as available for the reporting window.

2. Compare with the immediately preceding period of the same duration; also compare with the same weekday/window one week earlier when useful.

3. Look for changes and anomalies such as elevated error/warn log volume, new or noisy services, latency/throughput/resource changes, triggered monitors, unusual event spikes, or material drops in traffic.

4. Keep the Slack report short: title plus at most 5 bullets. Include key numbers/percent changes when available. If nothing material changed, say so briefly.

5. If Datadog has insufficient data or an API query fails, mention that limitation briefly rather than failing silently.

6. Post the report to Slack using chat.postMessage to channel C0AGCTEREE9. Include this disclosure at the end: "_This report was generated by an AI agent (OpenHands) on behalf of the requester._"