# Phase 9 Output Generation

> Phase 9 Output Generation guide with Threat Composer export reference. Use when generating final reports, exporting to JSON/Markdown, or completing the threat modeling process.

- Skill: `awslabs/phase-9-output-generation` (Agent Skill)
- Install (CLI): `npx skillmds@latest add awslabs/phase-9-output-generation`
- Raw SKILL.md: https://api.skillmd.com/api/skills/awslabs/phase-9-output-generation/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Security
- Author: awslabs (https://skillmd.com/u/awslabs)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/awslabs/phase-9-output-generation

---


# Phase 9: Output Generation and Documentation

## Objective
Generate final deliverables: Threat Composer-compatible JSON export and human-readable Markdown report. All files are saved to the `.threatmodel/` directory.

## Tools Reference

### export_threat_model
Exports the current model and includes a state summary in the response.
- Omit `output_path` to use a timestamped filename
- Pass `output_path="my_model.json"` to choose a base path and filename
- Generates BOTH `.tc.json` and `.md` files
- Saves both files to the adjacent `.threatmodel/` directory
- JSON is Threat Composer compatible (schema version 1)

The comprehensive export includes current threat and mitigation statuses.
Extended JSON and Markdown also include residual-risk assessments; the standard
Threat Composer fields remain unchanged.

### Progress Tools
- `manage_workflow(action="progress")` -- Final progress summary with phase completion
- `manage_assumptions(action="list")` -- All documented assumptions

## Output Files

### JSON Export (Threat Composer Compatible)
```
.threatmodel/comprehensive_threat_model_YYYYMMDD_HHMMSS.tc.json
```

**Schema** (version 1):
```json
{
  "schema": 1,
  "applicationInfo": { "name": "", "description": "" },
  "architecture": { "description": "" },
  "dataflow": { "description": "" },
  "assumptions": [...],
  "mitigations": [...],
  "assumptionLinks": [],
  "mitigationLinks": [...],
  "threats": [...]
}
```

### Threat Composer Field Constraints
| Field | Max Length |
|---|---|
| threatSource | 200 chars |
| prerequisites | 200 chars |
| threatAction | 200 chars |
| threatImpact | 200 chars |
| statement | 1400 chars |
| tags (each) | 30 chars |

### Threat Composer Status Values
| Threats | Mitigations |
|---|---|
| threatIdentified | mitigationIdentified |
| threatResolved | mitigationInProgress |
| threatResolvedNotUseful | mitigationResolved |
| | mitigationResolvedWillNotAction |

### Markdown Export (Human-Readable)
```
.threatmodel/comprehensive_threat_model_YYYYMMDD_HHMMSS.md
```

Contains:
- Executive summary with key statistics
- Business context and features
- System architecture (components, connections, data stores)
- Threat actors with relevance/priority
- Trust boundaries (zones, crossing points, boundaries)
- Assets and flows with risk levels
- Threats grouped by status
- Mitigations grouped by status with linked threats
- Code-validation findings and implementation evidence
- Assumptions with rationale
- Phase progress table

## Importing to AWS Threat Composer
1. Open AWS Threat Composer
2. Click Import
3. Select the `.tc.json` file from `.threatmodel/`
4. All threats, mitigations, assumptions, and links will load

## Workflow

1. **Call `manage_workflow(action="guidance", phase="9")`**
2. **Call `export_threat_model()`**, optionally with `output_path="my_model.json"`
3. **Call `manage_workflow(action="progress")`** for final summary
4. **Present to user**:
   - File locations in `.threatmodel/`
   - Summary: N threats, N mitigations, N assumptions
   - Overall completion percentage
   - Key findings and open risks

## Completion Criteria
- [ ] JSON export generated (Threat Composer compatible)
- [ ] Markdown report generated
- [ ] Files saved to `.threatmodel/` directory
- [ ] `manage_workflow(action="status")` reports Phase 9 complete for the current model
- [ ] Progress summary generated
- [ ] User informed of file locations and key findings