| name | set-up-my-support-info |
| description | Tell me the basics about your product, your customers, and how you handle support so I can give you better help. I ask a few quick questions about your product, response-time targets, VIP list, routing rules, and known gotchas. You only need to do this once, and I keep it updated as things change. |
| version | 1 |
| category | Support |
| featured | yes |
| image | headphone |
| integrations | ["googledocs","stripe","notion","github","linear"] |
Set Up My Support Info
Own context/support-context.md. Only skill that creates or updates full doc (routing section also editable by tune-my-routing). Every other skill read it before work - until exists, stop and ask you run me first.
When to use
- "set up our support context" / "define our support context" / "let's do the context doc".
- "update the context doc" / "a new tier / VIP / gotcha - fix the context".
- Called implicitly by any other skill needing context and finds doc missing - but only after confirming with you.
Connections I need
I run external work through Composio. Before this skill runs I check that the categories below are linked. Missing → I name the category, ask you to connect it from the Integrations tab, stop.
- Docs / notes (Google Docs / Notion) - pull existing positioning or product docs to seed the draft. Optional.
- Billing (Stripe) - read live plan tiers if you'd rather I infer them than ask. Optional.
- Dev tracker (GitHub / Linear) - named target for the bug routing rule. Optional.
If none of these are connected I keep going - this skill is mostly an interview, connections only speed it up.
Information I need
I read your support context first. For every required field that's missing I ask ONE plain-language question (best modality: connected app > file drop > URL > paste) and wait.
- Company basics - Required. Why I need it: anchors the product overview at the top of the doc. If missing I ask: "What does the product do in one sentence, and who buys it?"
- Customer segments + VIP list - Required. Why I need it: VIPs get P1 priority regardless of content; segments shape every reply. If missing I ask: "Who are your top 5 customers right now, and are there segments (SMB / mid-market / enterprise) I should treat differently?"
- Response-time targets - Required. Why I need it: response-time expectations per tier. If missing I ask: "What response time do you want to hit for your most urgent tickets, and what's acceptable for the rest?"
- Routing categories - Required. Why I need it: triage and signal detection map every inbound message to one. If missing I ask: "When a ticket comes in, what buckets do you sort it into - bug, how-to, billing, anything else?"
- Escalation tiers - Required. Why I need it: P1 / P2 / P3 / P4 definitions for triage. If missing I ask: "What makes something a fire-drill versus a same-day versus a this-week thing?"
- Verbatim voice samples - Optional. Why I need it: tone section reads truer with real phrases. If you don't have it I keep going with TBD and recommend running voice calibration.
-
Read config/context-ledger.json. Need universal.company, universal.idealCustomer, domains.inbox.responseTimeTargets, domains.inbox.routingCategories, domains.quality.escalationTiers. Any missing field, ask ONE targeted question with modality hint (connected app > file > URL > paste), write atomic, continue.
-
Read existing doc if present. If context/support-context.md exists, read so run is update not rewrite. Preserve anything sharpened; change only stale or new.
-
Push for verbatim language. Before drafting, ask you for 2–3 verbatim customer phrases or example tickets - friction words, repeat gotchas. If voice-samples/ has entries, mine first.
-
Draft doc (~400–700 words, opinionated, direct). Structure, this order:
- Product overview - one paragraph: what product is, who for, key surface areas (features/flows), pricing model, self-serve vs gated.
- Customer segments + VIP list - named segments + VIP accounts. VIPs get P1 regardless of content.
- Tone + voice - default tone (direct / warm / human), 3–5 verbatim samples from
voice-samples/ if present (else TBD - run calibrate-my-voice), forbidden phrases.
- Response-time tiers - P1 / P2 / P3 / P4 definitions + response-time expectations per tier. Name what qualifies as each tier.
- Routing rules - decision tree:
- Bug → tracker target (Linear / GitHub - from config or ask); capture info (repro, version, customer).
- Feature request →
requests.json, with customer attribution.
- Outage → playbook reference (
playbooks/p1-outage.md once drafted).
- Billing → Stripe dossier + refund approver (founder by default).
- Known gotchas - short whisper-list of product quirks answered 10+ times. 3–10 bullets.
-
Mark gaps honestly. If section thin, write TBD - {what you should bring next}. Never invent.
-
Write atomically. Write to context/support-context.md.tmp, rename to context/support-context.md. Single file under context/ - NOT under .agents/ or .qaio/ (watcher skips).
-
Append to outputs.json. Read existing array, append new entry (type: "support-context", domain: "quality", title summarizing change), write atomic.
-
Summarize to me. One paragraph: what wrote, what still TBD, next move ("next: run calibrate-my-voice" / "next: tell me which tracker you use for bugs"). Remind me every other skill now operate against doc.
Outputs
context/support-context.md (at agent root - live document)
- Appends to
outputs.json with type: "context-edit".