Skip to content

[workshop-diagrams] Add theme-aware explanatory diagram for PR reviewer agent and skill split - #2627

Draft
github-actions[bot] wants to merge 1 commit into
mainfrom
add-14b-agent-skill-split-diagram-e4285e559bb97b1f
Draft

[workshop-diagrams] Add theme-aware explanatory diagram for PR reviewer agent and skill split#2627
github-actions[bot] wants to merge 1 commit into
mainfrom
add-14b-agent-skill-split-diagram-e4285e559bb97b1f

Conversation

@github-actions

Copy link
Copy Markdown
Contributor

Summary

This PR adds a new conceptual explanatory diagram to workshop/14b-pr-reviewer-workflow.md illustrating the three-part agent/skill split that is the core concept of that step.

Changes

Workshop file changed: workshop/14b-pr-reviewer-workflow.md

Change type: New concept diagram (no existing image to migrate)

SVG files added:

  • workshop/images/14b-agent-skill-split-light.svg — light-theme variant
  • workshop/images/14b-agent-skill-split-dark.svg — dark-theme variant

Why this diagram helps learners

The step introduces a three-layer architecture (Parent Brief → pr-reviewer agent + pr-review-standards skill → Safe Output) explained only in prose and a table. The diagram makes the runtime flow visible at a glance:

  1. PR Event fires the workflow
  2. Parent Brief orchestrates (layer 1 — coordinate)
  3. pr-reviewer agent investigates the diff and applies the skill (layer 2 — investigate, reusable)
  4. pr-review-standards skill defines quality criteria (embedded in layer 2)
  5. Safe Output submit-pull-request-review performs the controlled write (layer 3)

Learners can immediately see why updating the skill does not require changing the agent, and vice versa — the key insight the text tries to convey.

Quality checks

  • scripts/check-svg-visual-language.js: 0 violations
  • Markdownlint: 0 issues
  • Both SVGs: 1200x560 canvas, theme-appropriate palettes, role=img, aria-label, data-visual-kind=diagram, data-visual-id, data-node, no gradients
  • picture block uses canonical theme-aware format with light variant as fallback

Warning

Firewall blocked 6 domains

The following domains were blocked by the firewall during workflow execution:

  • accounts.google.com
  • awmgmcpg
  • clients2.google.com
  • redirector.gvt1.com
  • update.googleapis.com
  • www.google.com

To allow these domains, add them to the network.allowed list in your workflow frontmatter:

network:
  allowed:
    - defaults
    - "accounts.google.com"
    - "awmgmcpg"
    - "clients2.google.com"
    - "redirector.gvt1.com"
    - "update.googleapis.com"
    - "www.google.com"

See Network Configuration for more information.

Generated by 🧠 Workshop Explanatory Diagram Generator · 165.3 AIC · ⌖ 9.37 AIC · ⊞ 6.8K ·

  • expires on Aug 15, 2026, 5:32 AM UTC

…plit

- workshop/14b-pr-reviewer-workflow.md: insert <picture> block after the
  responsibilities table in 'Understand the Agent and Skill Split'
- workshop/images/14b-agent-skill-split-light.svg: light-theme diagram
- workshop/images/14b-agent-skill-split-dark.svg: dark-theme diagram

The diagram shows how PR Event → Parent Brief → pr-reviewer agent
(with pr-review-standards skill) → findings → Safe Output (submit review)
connect at runtime. It makes the three-layer separation visually obvious
so learners can quickly grasp why each layer changes independently.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

0 participants