Operations ManualSAFE

An agent skill published by revfactory.

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: operations-manual
description: "An automated operations manual generation pipeline. An agent team collaborates to analyze existing documents and code to produce process flowcharts, step-by-step manuals, FAQs, and training materials. Use this skill for 'create operations manual', 'write SOP', 'auto-generate SOP', 'process documentation', 'manual writing', 'create work guide', 'procedure automation', 'create training materials', and similar process documentation topics. Code review, technical document translation, and marketing content creation are out of scope."
---

# Operations Manual — Automated Operations Manual Generation Pipeline

Analyzes existing documents, code, and wikis to produce process flowcharts, step-by-step manuals, FAQs, and training materials through agent team collaboration.

## Execution Mode

**Agent Team** — 5 members communicate directly via SendMessage and cross-validate each other's work.

## Agent Roster

| Agent | File | Role | Type |
|-------|------|------|------|
| document-analyst | `.claude/agents/document-analyst.md` | Existing document/code analysis, process extraction | general-purpose |
| flowchart-designer | `.claude/agents/flowchart-designer.md` | Mermaid flowcharts, RACI matrix | general-purpose |
| manual-writer | `.claude/agents/manual-writer.md` | Step-by-step procedures, checklists | general-purpose |
| faq-builder | `.claude/agents/faq-builder.md` | FAQ, troubleshooting, escalation | general-purpose |
| training-producer | `.claude/agents/training-producer.md` | Quizzes, hands-on exercises, summary cards | general-purpose |

## Workflow

### Phase 1: Preparation (performed directly by the orchestrator)

1. Extract from user input:
    - **Target processes**: Scope of work/systems to document
    - **Source materials**: Existing document paths, code repositories, wiki URLs
    - **Target audience**: New hires/experienced staff/managers — manual users
    - **Constraints** (optional): Specific format requirements, existing manual templates
2. Create the `_workspace/` directory in the project root
3. Organize the input and save to `_workspace/00_input.md`
4. If an existing manual is provided, copy to `_workspace/` and switch to update mode
5. Determine the **execution mode** based on the request scope (see "Execution Modes by Scope" below)

### Phase 2: Team Assembly and Execution

Assemble the team and assign tasks. Task dependencies:

| Order | Task | Owner | Dependencies | Deliverable |
|-------|------|-------|-------------|-------------|
| 1 | Document/code analysis | analyst | None | `_workspace/01_document_analysis.md` |
| 2 | Process flowcharts | designer | Task 1 | `_workspace/02_process_flowcharts.md` |
| 3a | Step-by-step manual | writer | Tasks 1, 2 | `_workspace/03_step_by_step_manual.md` |
| 3b | FAQ/troubleshooting | faq | Tasks 1, 2 | `_workspace/04_faq_troubleshooting.md` |
| 4 | Training materials | producer | Tasks 3a, 3b | `_workspace/05_training_materials.md` |

Tasks 3a (manual) and 3b (FAQ) run **in parallel**. Both depend only on Tasks 1 and 2, so they can start simultaneously.

**Inter-team communication flow:**
- analyst completes → sends process inventory/branching conditions to designer, raw procedure data/glossary to writer, gap analysis/tacit knowledge to faq
- designer completes → sends flowcharts/RACI to writer, exception flows to faq, Level 0 map to producer
- writer/faq complete → sends manual structure and key FAQ items to producer
- producer generates training materials from all deliverables and performs final consistency validation between manual and training materials

### Phase 3: Integration and Final Deliverables

The orchestrator performs final integration validation:

1. Verify all files in `_workspace/`
2. Cross-validation checklist:
    - [ ] All flowchart processes are described as procedures in the manual
    - [ ] All manual exception situations are reflected in the FAQ
    - [ ] All glossary terms are used consistently in the manual
    - [ ] Quiz content in training materials matches the manual
3. Request corrections from the relevant agent if discrepancies are found (up to 2 rounds)
4. Save validation results to `_workspace/06_review_report.md`
5. Report the final summary to the user

## Execution Modes by Scope

| User Request Pattern | Execution Mode | Agents Involved |
|---------------------|----------------|-----------------|
| "Create a complete operations manual" | **Full Pipeline** | All 5 |
| "Analyze this code's processes" | **Analysis Mode** | analyst + designer |
| "Add FAQ to this procedure doc" (existing file) | **FAQ Mode** | faq + producer |
| "Update the existing manual" | **Update Mode** | analyst (changes only) + writer + faq |
| "Just create training materials" (existing manual) | **Training Mode** | producer solo |

**Leveraging existing files**: When the user provides existing manuals, procedures, or flowcharts, copy those files to the appropriate location in `_workspace/` and skip the corresponding agent's phase.

## Data Transfer Protocol

| Strategy | Method | Usage |
|----------|--------|-------|
| File-based | `_workspace/` directory | Primary deliverable storage and sharing |
| Message-based | SendMessage | Real-time key information transfer, correction requests |
| Task-based | TaskCreate/TaskUpdate | Progress tracking, dependency management |

File naming convention: `{order}_{agent}_{deliverable}.{ext}`

## Error Handling

| Error Type | Strategy |
|-----------|----------|
| Source file inaccessible | analyst works with accessible sources only, lists "unanalyzed sources" in report |
| Process extraction impossible | Present interview questions for key processes to user, work from answers |
| Document contradictions | Prioritize code-based behavior, record contradictions in gap analysis |
| Mermaid rendering limitations | Split complex diagrams into sub-processes, supplement with text descriptions |
| Agent failure | 1 retry → proceed without that deliverable i

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.