Files
gh-gtmagents-gtm-agents-plu…/commands/monitor-loyalty.md
2025-11-29 18:30:55 +08:00

49 lines
3.1 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
---
name: monitor-loyalty
description: Evaluates loyalty program performance, member health, and experiment impact.
usage: /loyalty-lifecycle-orchestration:monitor-loyalty --window 30d --detail full --dimensions "tier,region"
---
# Command: monitor-loyalty
## Inputs
- **window** time range (7d, 30d, quarter) for analysis.
- **detail** summary | full report depth.
- **dimensions** optional breakdowns (tier, cohort, region, channel).
- **experiments** optional list of campaigns/tests to review.
- **alert_threshold** optional KPI threshold for escalations.
### GTM Agents Pattern & Plan Checklist
> Mirrors GTM Agents orchestrator blueprint @puerto/plugins/orchestrator/README.md#112-325.
- **Pattern selection**: Monitoring typically runs **pipeline** (data → diagnostics → experiments → action plan). If diagnostics and experiment review can run in parallel, document a **diamond** segment with merge gate in the plan header.
- **Plan schema**: Save `.claude/plans/plan-<timestamp>.json` capturing window, feeds, task IDs, dependency graph (BI, CRM, experimentation), error handling, and success metrics (retention, liability, engagement lift, alert remediation time).
- **Tool hooks**: Reference `docs/gtm-essentials.md` stack—Serena for metric definition diffs, Context7 for SOPs/dashboards, Sequential Thinking for retro cadence, Playwright for verifying embedded dashboards.
- **Guardrails**: Default retry limit = 2 for failed data pulls or chart renders; escalation ladder = Member Insights Analyst → Loyalty Strategist → Exec sponsor.
- **Review**: Run `docs/usage-guide.md#orchestration-best-practices-puerto-parity` before distributing outputs to ensure dependencies + approvals captured.
## Workflow
1. **Data Pull** collect enrollment, engagement, redemption, CSAT, and revenue metrics by selected window.
2. **Health Diagnostics** evaluate tier mix, churn risk, point liability, and reward breakage.
3. **Member Insights** surface segments needing nudges, upgrade opportunities, or win-back plays.
4. **Experiment Review** summarize test results with lift, guardrails, and recommendations.
5. **Action Plan** produce prioritized backlog (journey tweaks, rewards refresh, messaging updates).
## Outputs
- Dashboard snapshot with KPI trends and drilldowns.
- Risk/Opportunity list referencing data points and owners.
- Experiment readouts plus next-test suggestions.
- Plan JSON entry stored/updated in `.claude/plans` for audit trail.
## Agent/Skill Invocations
- `member-insights-analyst` owns analysis + insights.
- `loyalty-strategist` validates roadmap adjustments.
- `member-insights` skill ensures segmentation + reporting standards.
## GTM Agents Safeguards
- **Fallback agents**: document substitutes (e.g., Loyalty Strategist covering Analyst) when analysts unavailable.
- **Escalation triggers**: if alert_threshold is breached twice or liability spikes, trigger GTM Agents rip-cord escalation and record remediation in plan JSON.
- **Plan maintenance**: update plan JSON/change log when metrics, thresholds, or tooling stack shift to keep audit history accurate.
---