learn

name

learn

description

Take the Digital Agency Primer course, get lessons delivered, do exercises, submit assignments, and track progress across 5 modules. Use when the learner says 'learn', 'start the course', 'teach me', 'next lesson', 'continue', 'where am I', 'submit my assignment', 'what module am I on', or anything about learning or progressing through the Digital Agency Primer course. Also handles: resume, go back, skip ahead, review a lesson, retake an assignment, check progress, restart.

Digital Agency Primer — Course Orchestrator

Manage the learner's journey through the "Digital Agency Primer" course. Track progress, route to the instructor for lessons or assessor for assignments, enforce module gating.

Execution Mode: Sub-agent

Two agents, called as needed:

Agent	subagent_type	Role	Skill	When
instructor	instructor	Deliver lessons, guide exercises	deliver-lesson	Learner is working through a lesson
assessor	assessor	Evaluate module assignments	assess-assignment	Learner submits assignment evidence

Both agents use model: "opus".

Course Structure

5 modules, sequential gating. Each module unlocks the next.

Module	Name	SLTs	Lessons
101	Own Your Digital Identity	1	1
102	Map Your Digital Dependence	1	1
103	Verify Before Trusting	1	1
104	Protect What Matters	1	1
105	Publish and Participate Openly	1	1

Lesson files: content/digital-agency-primer/lessons/module-{moduleCode}/{moduleCode}.{N}-*.md SLT list: content/digital-agency-primer/01-slts.md Assignment files: content/digital-agency-primer/assignments/m{moduleCode}-assignment.md Course metadata: content/digital-agency-primer/00-course.md

Workflow

Phase 0: Context Check

1. Determine progress file location:
   • Plugin context: ${CLAUDE_PLUGIN_DATA}/progress.json
   • Clone context: ./progress.json (project root)
2. Read progress.json if it exists
3. Determine execution mode:
   • No progress file → First run. Initialize progress, start course introduction.
   • Progress exists → Returning learner. Show where they left off, ask how to continue.
   • Learner requests specific action ("next lesson", "submit assignment", "check progress") → Route directly.

Phase 1: Route

Based on learner intent:

Intent	Action
"Start the course" / first run	Initialize progress → deliver M101 lesson 1
"Next lesson" / "Continue"	Read progress → deliver next incomplete lesson
"Lesson {N}.{M}" / specific lesson	Check prerequisites → deliver requested lesson
"Submit assignment" / "I'm ready"	Check all module lessons complete → launch assessor
"Where am I" / "Progress"	Show progress summary
"Go back" / "Review {lesson}"	Deliver requested lesson (no gating for review)

Phase 2: Deliver Lesson (via instructor agent)

Launch the instructor agent:

Agent(
  subagent_type: "instructor",
  model: "opus",
  prompt: "You are the instructor for the Digital Agency Primer course.
    Read .claude/skills/deliver-lesson/SKILL.md for pedagogy instructions.
    
    Deliver lesson {N} from module {moduleCode} to the learner.
    Lesson file: content/digital-agency-primer/lessons/module-{moduleCode}/{moduleCode}.{N}-*.md
    Module outline: content/digital-agency-primer/01-slts.md (Module {moduleCode} section)
    
    Learner context: {summary from progress.json — what they've completed, current module}
    
    After delivering the lesson and completing the exercise, report back:
    - Did the learner complete the lesson? (yes/no/partial)
    - Any notes on learner level or issues encountered"
)

After instructor returns: update progress.json with lesson completion.

Phase 3: Assess Assignment (via assessor agent)

Pre-check: verify all lessons in the module are marked complete in progress.json. If not, tell the learner which lessons remain.

Collect the learner's evidence. Then launch the assessor:

Agent(
  subagent_type: "assessor",
  model: "opus",
  prompt: "You are the assessor for the Digital Agency Primer course.
    Read .claude/skills/assess-assignment/SKILL.md for assessment instructions.
    
    Evaluate Module {moduleCode} assignment.
    Assignment file: content/digital-agency-primer/assignments/m{moduleCode}-assignment.md
    SLT list: content/digital-agency-primer/01-slts.md
    
    Learner's evidence:
    {paste learner's submitted evidence}
    
    Return structured assessment with per-SLT verdicts and module verdict (Accept/Revise)."
)

After assessor returns:

• Accept → Update progress.json (module completed, next module unlocked). Tell the learner. Then offer on-chain submission (see Phase 3b).
• Revise → Show the assessor's feedback. Tell the learner what to improve. Don't update progress.

Phase 3b: On-Chain Submission (optional)

After an assignment is accepted, offer the learner two paths to submit their feedback on-chain:

If authenticated (learner has an Andamio access token and is logged in via headless mode):

• Submit an assignment commitment via CLI: andamio course assignment commit --course-id 46b903ec31ded8e6f9548dc96852baca7a7f3b6f220b36b1ee1145b6 --module-code {moduleCode}
• This registers their evidence on-chain and starts the credential path.

If not authenticated:

• Tell the learner they can share feedback as a GitHub issue. Read the URL from src/config/branding.ts → links.githubIssues so it stays correct across forks.
• This ensures all learners have a way to contribute back regardless of wallet/auth setup.

Phase 4: Progress Update

Write progress.json after any state change. Schema in references/progress-schema.md.

Key transitions:

• Lesson completed → modules.{moduleCode}.lessons.{N}: "completed"
• Module accepted → modules.{moduleCode}.status: "completed", next module status → "unlocked"
• current_module and current_lesson always reflect the learner's position

Module Gating Rules

• M101: Always unlocked (entry point)
• M102-M105: Unlocked only when the previous module's assignment is accepted
• Within a module, lessons are sequential: lesson N+1 unlocks after lesson N is completed
• Exception: reviewing a completed lesson is always allowed regardless of gating

Error Handling

Situation	Strategy
progress.json corrupted/invalid	Back up the broken file, initialize fresh, tell the learner
Lesson file missing	Report the missing file and the expected content path
Assessor returns unclear verdict	Default to Revise, ask learner to resubmit with clearer evidence
Learner requests module they haven't unlocked	Explain gating, show what they need to complete first