FedEx Ship Manager · Hoofddorp Knowledge Base Agent · Copilot CLI

Magic Agent

The MAGIC+ Knowledge Base Agent for FedEx Ship Manager — a super-powered AI agent with endless memory and access to all MAGIC tools. 12 specialized skills, 70+ domain docs, and the ability to debug, refine stories, and keep its own knowledge current.

Officially adopted across 7+ teams via Copilot CLI — QAs, developers, and Product Owners use it daily to answer questions, debug issues, and ship faster, straight from the terminal.

Copilot CLI Agent Skills RAG Dynatrace PostgreSQL Playwright Figma

12 AI Skills

7+ Teams

70+ Domain Docs

Architecture

Endless memory, 12 specialized skills, and live access to the team's real tools — all driven from the terminal.

Knowledge Base

70+ domain docs

ShipmentsRatesCustomsBillingReturnsHazmat

Magic Agent

12 specialized skills · endless memory

Connected Tools

DynatraceOptimizelyFigmaVersionOnePostgreSQLMAGIC APIs

Copilot CLI · Terminal

QAs · Developers · Product Owners

Endless Memory & Real Tool Access

70+ domain docs cover every aspect of Ship Manager — shipments, rates, customs, billing, returns, hazmat, and more — and the agent connects to the real tools the team uses every day.

Dynatrace

Optimizely

Figma

VersionOne

PostgreSQL

MAGIC APIs

12 Specialized Skills

Each skill is an expert in its domain — the agent picks the right one automatically. A few of the ones that move the needle most:

/magic

Ask any Ship Manager business or technical question. The agent pulls the most relevant context from 70+ domain docs and answers with precision.

/dynatrace-logs

Debug a production issue end-to-end: reads Dynatrace logs, checks Optimizely feature toggles, traces API calls, and pinpoints the root cause — in ~3 minutes instead of hours of manual replication.

/refinement {V1}

Turns a VersionOne story ID into a complete refinement report: scope and impacted areas, cross-team dependencies and risks, and acceptance-criteria gaps — pulling from knowledge and Figma.

/update-knowledge

One command scans merged PRs and V1 stories across repos, reads the diffs, updates the matching domain docs, and generates an HTML report — keeping the knowledge base current without manual work.

/playwright-expert

Reads a pull request, understands the changes, and generates Playwright tests following the team's best practices and page object patterns.

/api-caller

Generates ready-to-run API calls and Postman collections with pre-filled test data — straight from a PR or OpenAPI spec.

What's Next — From Tech Teams to Everyone

Today the agent runs in the terminal via Copilot CLI for technical teams. The same knowledge base and skills can go online in the browser for every employee — and, with legal and security approval, power a customer-facing agent:

Shipment Tracking

"What was my last shipment?" — the agent looks it up, shares status, and offers to navigate you straight to it.

Error Resolution

"Why do I see this error?" with a screenshot — the agent explains the cause (e.g. customs needs an email) and how to clear it.

Smart Guidance

"Ship 2 packages, 30KG, from US to NL" — the agent gathers the needed info and estimates cost and delivery.

Auto-Fill from Invoice

Upload an invoice and the agent extracts commodity, HS code, dimensions, and weight — ready to review and submit.

Same knowledge base, same skills, same agent. Going online for non-technical teams and customers needs legal review, security approval, and data-privacy assessment.

Want to know more about Magic Agent?

Let's Talk