Use case · Just you · Specialist skills
Reply with built-in legal, inventory, and pricing skills
You don't have a legal team. You don't have an inventory analyst. You don't have a pricing specialist. Magnis runs each as a skill on your graph — answers in your inbox, with verified output. The team you didn't hire.
Demo workspace — Legal, Inventory, and Pricing skills run on simulated business data (contract templates v3.2, EU warehouse stock log, tier-2 pricing rules). In v1, skills are role-prompted single-user. Shared team graph with skill-permissions on the roadmap.
Step 1
Describe the situation
A customer email lands. It needs three different specialists to answer: someone who knows the warehouse, someone who knows the contract clauses, someone who knows the pricing rules. You don't have any of them.
Customer email — incoming
Hi — quick question. We need 12 units of part X delivered before Friday. Also, our procurement requires a standard liability clause in the SOW. And — given we ordered $40K with you last quarter, do we still get the volume discount?
Step 2
Skills sit in your right rail
Each skill is a role-scoped agent with its own prompt and source access. They run on the same graph, but answer only in their lane.
Legal-bot
Reviews clauses + flags deprecated language
Inventory-bot
Live stock check across warehouses
Pricing-bot
Discount tier rules + customer eligibility
HR-bot
Employment + contractor onboarding policy
Step 3
The consultation
You drive. Each skill answers in its lane with provenance. Reply assembles in your composer with all three inputs verified.
You
Draft a reply confirming we can deliver 12 units of part X by Friday.
Inventory-bot
✅ 12 units in EU warehouse — Berlin DC. Lead time 3 days standard, 1 day expedited (+€80). Stock confirmed at 09:14 today.
Source · EU warehouse stock log
You
Add the standard liability clause to the SOW.
Legal-bot
Use clause v3.2 — v3.1 has been deprecated since 2026-03 (regulatory update). Clause v3.2 covers the standard SOW liability cap with enforceable jurisdiction language.
Source · Contract templates · v3.2
You
What's the current discount tier for them — they ordered $40K last quarter?
Pricing-bot
Volume discount tier 2 applies (10% off list, $40K+ in trailing quarter). They qualify. Tier expires 2026-06-30 — flag if order doesn't repeat.
Source · Pricing rules · tier 2
You
Looks good — assemble the reply.
Magnis
Drafted reply ready in your composer. Confirms 12 units / Friday delivery via Berlin DC, attaches SOW with clause v3.2, applies 10% volume discount with tier-2 footnote. All claims linked to source records (warehouse log + contract template + pricing rules). Awaiting your approval to send.
Step 4
Follow-up prompts
Add HR-bot for hiring questions
“Onboard a contractor with NDA terms — HR-bot drafts the engagement letter, Legal-bot reviews.”
Skill stack per inbox lane
“When a customer email mentions pricing, auto-consult Pricing-bot before draft.”
Audit trail in graph
“Show me every customer reply where Legal-bot was consulted in the last quarter.”
Step 5
Tips & troubleshooting
Skills aren't replacements for senior judgement.
Legal-bot covers ~95% of routine clauses. For the 5% — bet-the- company terms, novel jurisdictions — bring a real lawyer. Magnis flags low-confidence answers explicitly.
Provenance is the trust gate.
Every skill answer links to its source — clause version, warehouse record, pricing rule version. Click through before shipping the reply. No black-box claims.
Built-in vs custom skills.
v1 ships with built-in Legal / Inventory / Pricing / HR. You can define custom skills (your own role prompt + source bindings). What we don't have: a 3rd-party skill marketplace — that's on the roadmap.
The team you don't need to hire.
Connect Magnis. Activate skills you need. Reply to customers with specialist input — your data, your machine, no headcount.