How-To Guides: Case Management

Problem-oriented guides for common case management scenarios.

1. Creating a New Case

Problem: Need to start a knowledge work project

Solution:

1. Define desired outcome:

   1
   2
   "What does success look like when this case completes?"
   Example: "Comparative ML framework analysis with deployment recommendation"
   

2. Select template:
   • Research: Literature-driven investigation
   • Development: Software/product creation
   • Consulting: Client engagement
   • Creative: Original content creation
3. Configure case:

   1
   2
   3
   4
   5
   case:
     desired_outcome: "Your outcome here"
     template: research_case_pattern
     owner: your-id
     members: [human-001, ai-specialist-g3]
   

4. Activate:

   1
   sea case create --config case-config.yaml
   

2. Defining Custom Sentries

Problem: Need stage-specific activation conditions

Solution:

Entry Sentry Example:

1
2
3
4
5
6
7
8
sentry:
  type: entry
  stage: analysis
  conditions:
    - data_collected: true
    - quality_score: >= 0.8
    - peer_review: passed
  logic: AND  # All conditions must be true

Exit Sentry Example:

1
2
3
4
5
6
7
8
sentry:
  type: exit
  stage: analysis
  conditions:
    - findings_documented: true
    - statistical_significance: >= 0.95
    - stakeholder_review: approved
  logic: AND

Discretionary Task Activation:

1
2
3
4
5
6
7
sentry:
  type: task_activation
  task: additional_data_collection
  conditions:
    - sample_size: < 100
    - confidence_interval: > 0.1
  logic: OR  # Any condition triggers

3. Managing Artifact Pipeline

Problem: Artifact progression unclear

Solution:

Step 1: Identify Current Stage

1
2
sea artifact show art-hypothesis-doc
# Output: Stage: Cognitive

Step 2: Check Transition Requirements

1
2
3
4
Cognitive → Intellectual requires:
- Synthesis task complete
- Quality review passed
- TransitionToken generated

Step 3: Execute Transition

1
2
3
sea artifact synthesize art-hypothesis-doc --to intellectual
# Generates: tt-synthesis-001
# Creates: Intellectual Artifact

Step 4: Verify Lineage

1
2
sea artifact lineage art-hypothesis-doc
# Shows: Cognitive → Intellectual (tt-synthesis-001)

4. Configuring PM-Agent

Problem: Need custom task assignment logic

Solution:

Step 1: Define Assignment Rules

1
2
3
4
5
6
7
8
9
10
11
12
13
pm_agent:
  assignment_rules:
    - rule: capability_match
      priority: 1
      logic: "Match task requirements to member capabilities"
    
    - rule: workload_balance
      priority: 2
      logic: "Distribute tasks evenly across members"
    
    - rule: authority_check
      priority: 3
      logic: "Verify SDS-031 privileged action rules"

Step 2: Configure Knowledge Base

1
2
3
4
5
6
7
pm_agent:
  knowledge_base:
    context: ml-frameworks-evaluation
    sources:
      - domain_concepts
      - past_cases
      - member_profiles

Step 3: Set Sentry Evaluation

1
2
3
4
5
pm_agent:
  sentry_evaluation:
    frequency: on_task_complete
    auto_activate: true
    log_decisions: true

Step 4: Apply Configuration

1
sea pm-agent config --file pm-agent-config.yaml

5. Handling Case Suspension

Problem: Need to pause case temporarily

Solution:

Suspend Case:

1
sea case suspend case-research-001 --reason "Awaiting external data"

Effects:

• All active tasks paused
• PM-Agent stops new assignments
• Artifacts preserved in current state
• TransitionTokens logged

Resume Case:

1
sea case resume case-research-001

Effects:

• Tasks reactivate
• PM-Agent resumes orchestration
• Sentries re-evaluated
• Work continues from suspension point

6. Creating Discretionary Tasks

Problem: Need optional tasks that activate conditionally

Solution:

Define Discretionary Task:

1
2
3
4
5
6
7
8
task:
  id: task-additional-analysis
  type: discretionary
  activation_sentry:
    conditions:
      - initial_results: inconclusive
      - confidence: < 0.8
  assignee: auto  # PM-Agent assigns

PM-Agent Evaluation:

1
2
3
4
5
1. Monitor sentry conditions
2. When conditions satisfied:
   - Activate task
   - Assign to qualified member
   - Log activation to IFL

Manual Activation (Override):

1
2
sea task activate task-additional-analysis --reason "Stakeholder request"
# Logged as manual override

7. Capitalizing Artifacts

Problem: Convert work product to organizational asset

Solution:

Step 1: Identify Capital Candidate

1
2
3
4
Criteria:
- Reusable across projects
- Proven methodology
- Generalizable knowledge

Step 2: Prepare for Capitalization

1
2
3
4
5
6
# Verify artifact is Information Product
sea artifact show art-methodology

# Check Exchange Rate Policy
sea capital policy show
# Example: 40 hours work → $8,000 capital value

Step 3: Request Approval (SovereignGate)

1
2
sea artifact capitalize art-methodology --approver human-sovereign
# Requires: SignatureToken from human approver

Step 4: Complete Capitalization

1
2
3
4
5
Upon approval:
- Artifact promoted to Capital stage
- TransitionToken generated
- Capital value assigned
- Added to organizational knowledge base

8. Troubleshooting

Issue: Sentry Won’t Activate

Diagnosis:

1
2
sea sentry debug sentry-analysis-entry
# Shows: Which conditions failed

Solution:

1. Review failed conditions
2. Complete missing tasks
3. Generate required artifacts
4. Re-evaluate sentry

Issue: PM-Agent Not Assigning

Diagnosis:

1
2
sea pm-agent logs --case case-research-001
# Shows: Assignment decision log

Solution:

1. Check member capabilities match task requirements
2. Verify workload capacity
3. Review authority rules (SDS-031)
4. Manual override if justified

Issue: Artifact Pipeline Broken

Diagnosis:

1
2
sea artifact lineage art-product-001
# Shows: Missing TransitionToken

Solution:

1. Identify missing stage
2. Create required artifact
3. Generate TransitionToken
4. Verify lineage complete