Customer Support

• You're launching a new product or service and need a customer support system built from scratch.
• You have a fragmented CS setup (scattered FAQs, inconsistent response scripts) and want a structured, consolidated system.
• You need a specific deliverable — just an FAQ, just an escalation policy, just a response manual — without running the full pipeline.
• You have existing CS materials (an FAQ file, old scripts) and want them reviewed, extended, or complemented with missing components.

• You need CRM or helpdesk software developed (e.g., custom Zendesk integrations, ticketing systems) — this skill produces documentation, not code.
• You need a real-time chatbot built and deployed — this skill creates the content a chatbot would use, not the chatbot itself.
• You need call center infrastructure setup or agent recruitment support — those are outside the scope of this skill.

A single prompt triggers parallel and sequential generation of FAQ, response manual, escalation policy, and CS analytics framework. All four deliverables are saved as structured files in a _workspace/ directory and cross-validated before delivery.

Dedicated agents handle FAQ construction, response scripting, escalation tier design, metrics/VOC analysis, and final review. The reviewer agent simulates customer journeys and flags critical issues, sending revision requests back to the relevant agent (up to two rounds) before finalizing.

Not every request needs the full pipeline. If you only need FAQs, a response manual, an escalation policy, analytics design, or a review of existing materials, the skill deploys only the agents relevant to that task — keeping output focused and efficient.

If you already have an FAQ, manual, or data file, you can attach it and the pipeline will copy it into the workspace and skip the corresponding generation step, building only what's missing.

The escalation-manager agent produces a tiered L1/L2/L3 escalation structure, severity classification, trigger conditions, SLA matrix, and crisis response protocol — informed directly by the outputs of the FAQ and response manual agents.

The cs-analyst agent designs a metrics framework covering CSAT, NPS, FCR, and AHT, along with weekly and monthly reporting templates. Targets are set using industry benchmarks when no existing data is available.

A B2C SaaS company with 10,000 MAUs and a 5-person support team needs a full CS system for chat and email channels. The full pipeline produces 30+ FAQ Q&As across billing/features/integration categories, channel-specific response scripts, an L1–L3 escalation matrix, and a CSAT/NPS dashboard template.

A team already has an FAQ document but lacks response scripts and escalation procedures. By attaching the existing FAQ file, the pipeline skips FAQ generation and deploys only the response-specialist, escalation-manager, cs-analyst, and reviewer agents to fill the gaps.

A support team experiencing increasing ticket volume needs a formal escalation policy before hiring L2 agents. Escalation Mode deploys just the escalation-manager and reviewer agents to produce a tiered escalation structure with SLA definitions and crisis response protocols.

An operations manager wants an independent review of their existing FAQ and response manual for consistency, coverage gaps, and customer journey alignment. Review Mode deploys the cs-reviewer agent alone to simulate customer journeys and produce a structured review report.