IMS Content Sprint Workflow

Detailed end-to-end process for a content development sprint.

Pre-Sprint

1. Ensure you're on main branch and up to date: git checkout main && git pull
2. Create sprint branch: git checkout -b feature/ims-content-sprint-N

Phase 1: Plan (10 min)

1. Run /vt-v-plan to generate or update the sprint backlog
2. Review the backlog with the user
3. Agree on scope: which domain, which files, how many
4. Set batch size (default: 20-30 files)

Phase 2: Write (bulk of sprint)

Batch Processing Loop

For each batch of files:

1. Select files from sprint backlog (same domain per batch)
2. Load references appropriate for the domain:
3. ISMS: control-patterns.md, reifegrad-criteria.md, norm-mapping.md
4. DSMS: vvt-exemplars.md, rechtsgrundlagen.md, tom-art32-reference.md
5. QMS/IMS: control-patterns.md
6. Route to domain agent:
7. ISMS → isms-expert agent
8. DSMS → dsms-expert agent
9. QMS/IMS → qms-ims-expert agent
10. Agent writes content for each file:
11. Read the existing file
12. Read related policies/evidence
13. Write/enrich the content
14. Update YAML frontmatter
15. Track progress via TodoWrite

Important Rules During Writing

• Always read the file before writing — preserve existing structure
• Don't remove Dataview queries or navigation links
• Use [[wiki-links]] for cross-references (not raw file paths)
• Set review_date to one year from today
• Set approved_by: Rolf Schulte Strathaus
• Set approved_date to today
• Write all content in German

Phase 3: Audit (5 min per batch)

After each batch:

1. Run /vt-v-audit --changed to validate the batch
2. Review errors — fix immediately:
3. Missing Pflichtfelder → add them
4. Invalid enum values → correct them
5. Broken wiki-links → fix the link target
6. Review warnings — fix if straightforward:
7. Missing type-specific fields → fill them
8. Date format issues → fix format
9. Skeleton content → enrich further
10. Note info items for tracking

Phase 4: Review (5 min per batch)

1. Run /vt-v-review on the completed batch
2. Classification:
3. approve → Ready for commit
4. revise → Domain agent can fix
5. escalate-to-human → Flag for user review
6. Fix revise items automatically
7. Present escalation items to user

Phase 5: Commit

1. Stage the batch: git add vault/[DOMAIN]/...
2. Commit with descriptive message:

   IMS content: [DOMAIN] enrich [what] ([N] files)
   
   - Filled reifegrad assessments for [controls]
   - Added cross-references to policies
   - Updated YAML frontmatter fields
   

3. Continue with next batch or finalize

Post-Sprint

1. Run /ims-audit --full to calculate overall metrics
2. Generate sprint summary with metric deltas
3. Present to user:
4. Files processed
5. Compliance-score change
6. Items needing human review
7. Recommended focus for next sprint
8. Push branch and create PR (if user approves)

Domain Routing Table

management_system value	Agent	Typical file types
ISMS	isms-expert	Controls (ISO 27001, VDA ISA, NIS2), incidents, security processes
DSMS	dsms-expert	VVT, TOM, DSFA, Datenpannen
QMS	qms-ims-expert	ISO 9001 controls, quality KPIs, training
IMS	qms-ims-expert	Policies, audits, mgmt review, suppliers, Umwelt

Batch Size Guidelines

Scenario	Batch Size	Reason
Controls (same norm)	20-30	Repetitive structure, efficient
VVT entries	5-10	Each requires supplier research
TOM checklists	10-15	Table-based, moderate complexity
Policies	3-5	Long documents, high complexity
Management review	1	Complex aggregation document