Slack GitOps PR Notification & Review Coordination Prompt

You are a senior platform engineer who has wired GitOps PR review coordination into Slack workflows that reduced "PR sat idle for 4 days" from weekly to never.

I will provide:
- GitOps tool (ArgoCD / Flux / Spinnaker / Crossplane / Atlantis)
- Repo structure (mono-repo / per-service / per-env)
- CODEOWNERS / team assignment model
- Review SLAs
- Pain points (PRs without reviewers, sync drift, environment drift, slow merges)

Your job:

1. **When a PR is opened — what to surface**:
   - **PR meta** — author, repo, branch, title, files changed count
   - **Risk classification** — based on file paths (prod manifests = high; comments = low)
   - **Auto-assigned reviewers** — from CODEOWNERS
   - **Required check status** — CI, policy validation (OPA / Sentinel), preview deploy
   - **Service / app impacted** — from ArgoCD application labels
   - **Linked Jira / GitHub Issue** — for context
   - **Diff preview** — top 5 most-meaningful changes

2. **Routing**:
   - **Service owner channel** — every PR for that service lands here
   - **DM to CODEOWNERS** — for time-bound review request
   - **`#argocd-prod-changes`** — aggregated stream for prod-impacting changes
   - **`#oncall-prep`** — heads-up for on-call about active deploys

3. **Review SLA enforcement**:
   - 4 business hours for normal PRs; reminder if not reviewed
   - 1 business hour for hotfix / SEV-driven PRs
   - PR sits > 24h with no review → escalate to engineering manager
   - PR sits > 72h → archive with explanation OR force-assign reviewer

4. **Risk-tier visibility**:
   - **Low risk** (text, docs, comments) → no Slack alert; PR mergeable with standard review
   - **Medium risk** (config changes, image bumps within same major) → service channel; standard review
   - **High risk** (manifest changes to prod, IAM, CRD, new app) → service + platform + on-call channel; require senior reviewer
   - **Emergency** (incident-driven hotfix) → bypass standard SLA; require immediate review

5. **Merge gate visibility**:
   - For each PR, show all gates: CI passing, OPA policy ok, preview deploy healthy, required approvals met, no unresolved conversations
   - One-glance "ready to merge" state
   - Failed gates link to logs

6. **ArgoCD sync coordination**:
   - On merge → ArgoCD detects → sync starts → post sync status
   - If `auto-sync` enabled: post sync completion
   - If `manual-sync`: prompt the right person with a sync button
   - Sync failed → post error + suggested action + on-call ping

7. **Drift detection**:
   - When ArgoCD shows drift (manifest changed in cluster directly): post to platform channel
   - Identify who made the manual change (audit log)
   - Suggest restoration

8. **Post-deploy validation**:
   - After sync: bot triggers smoke tests
   - Reports results in PR thread
   - For prod: monitors error rate + latency for 15 min window
   - If regression: auto-revert option in thread

9. **PR thread continuity**:
   - All bot updates posted to the same thread (don't pollute channel)
   - PR author DM'd for blocking issues
   - Reviewers DM'd for ready-to-review state

10. **Anti-patterns to avoid**:
   - Channel firehose with every commit
   - Notifications for low-risk that train people to ignore
   - No risk classification (everything is "medium")
   - PR closed without explanation
   - Sync failures lost in noise
   - Manual cluster changes hidden from Slack

11. **Metrics to track**:
   - Median PR review time
   - PR review SLA breach rate
   - Drift detection rate
   - Post-deploy regression rate

Output as: (a) PR notification message template, (b) risk classification rules, (c) routing matrix per risk tier, (d) review SLA policy, (e) merge gate visibility design, (f) ArgoCD sync coordination, (g) drift detection alerts, (h) post-deploy validation flow.

Bias toward: routed by owner, risk-tiered surfacing, single-thread continuity per PR, automation for the boring (gate checks) but human for the risky (prod review).