UiPath Coding Agents Workshop — Site

Hands-on course teaching clients & partners to build UiPath agents and Maestro orchestrations with coding agents (Claude Code) and the uip CLI.

Built on the same MkDocs framework as the Agentic Practice Workshop, with three additions: a private knowledge base, a configurable training environment, and a CLI/KB-driven lesson-generation path alongside the screenshot path.

../knowledge-base/  (private SSOT, NOT in this repo)
        ↓  used as context to generate accurate lessons
docs/*.md  →  MkDocs builds HTML  →  GitHub Actions  →  gh-pages  →  GitHub Pages

The knowledge base is intentionally kept outside this public repo (../knowledge-base/) so internal material is never published.

# 1. Pull the reusable framework files from the base repo
./bootstrap-framework.sh

# 2. Install dependencies
pip install -r requirements.txt

# 3. Preview locally (authoring environment + WIP nav)
COURSE_ENV=staging mkdocs serve -f mkdocs.local.yml
# open http://127.0.0.1:8000

# 4. Release-equivalent build check
COURSE_ENV=prod mkdocs build

This repo is scaffolded locally. To create the remote and deploy:

cd courses/coding-agents/site
git init && git add . && git commit -m "Initial scaffold"
gh repo create uipath-practice/CodingAgentsCourse --public --source=. --push
# then enable GitHub Pages on the gh-pages branch (created by the deploy workflow)

This section is for trainers/facilitators only. It stays in this README rather than docs/ because GitHub Pages has no way to restrict access by email — anything admin-only needs to live in the repo itself, not on the public site. Do not move this content into docs/ or reference it from mkdocs.yml/mkdocs.local.yml.

The site's main.py defines a COURSE_ENV switch with two profiles (staging and prod) driving the {{ training_url }} / {{ training_tenant }} macros used throughout the lessons. In practice, only the cloud environment below is real — there is no separate staging UiPath org/tenant to author or test against. Treat the staging profile in main.py as unused/vestigial.

All exercises operate inside a single shared Orchestrator folder in this org/tenant (tpenlabs / CodingAgentsPractice):

• Folder name: CodingAgentsILT
• Folder key: c30345cd-5543-46a9-b42b-0354e60b4f15

Every deployable artifact (agent, coded app, Maestro flow, RPA process) is named with a {YourName} prefix (e.g. {YourName}LoanUnderwritingAgent) so participants sharing the same tenant/folder don't collide with each other's packages, processes, or Orchestrator entities.

Before a cohort starts, invite each participant to the tpenlabs org so they can log in with their own email:

1. In the tpenlabs org, go to the Admin tab.
2. Go to Accounts and Local Groups.
3. Invite each participant using their email address.
4. Make sure each invited participant is added to the CodingAgentsGroup local group — this is what grants them access to the CodingAgentsPractice tenant and the CodingAgentsILT folder.

• Confirm the LoanUnderwriting storage bucket exists in the CodingAgentsILT folder in Orchestrator — the RPA lesson (Application Intake & Enrichment) uploads results there, and the run fails if it's missing.

• Is there a reset/cleanup step needed between cohorts (stale processes, packages, jobs, Action Center tasks left in the shared folder from the previous run)? Document the steps or the script here.
• Any UiBank API seed-data reset needed, or is https://uibank-api.uipath.com stable/shared indefinitely across cohorts?
• Who to contact if a participant needs their credentials/tenant access re-issued mid-workshop?

Not yet documented — add suggested pacing per lesson here.

Not yet documented — add reference solutions / expected results here.