Project Backup & Restore
Export complete project state to JSON, archive to S3, and restore from any snapshot. Full data portability with zero data loss.
How it works
1
Full project export
Export everything: project state, features, requests, events, execution steps, orchestration runs, and interaction sessions into a single JSON document.
2
S3 archival
Archive exports to MinIO/S3 with timestamped keys. Backups are stored in the iter-pipeline-artifacts bucket under {project_id}/backups/.
3
Automatic run archival
After every successful orchestration, completed runs are automatically archived to S3. Each run is stored individually under {project_id}/runs/{request_id}/.
4
Restore from any snapshot
Browse available archives, pick a timestamp, and restore. Import supports merge mode to layer data without overwriting existing state.
Technical details
API Endpoints
POST /projects/{id}/exportFull project exportPOST /projects/{id}/importImport from JSONPOST /projects/{id}/archiveExport + upload to S3GET /projects/{id}/archivesList S3 snapshotsPOST /projects/{id}/restore/{key}Restore from S3
What's exported
- Project state (features, requests, notes, assets)
- Event log (structured events with timestamps)
- Execution steps (pipeline step trees)
- Orchestration runs (configs, steps, reviews)
- Interaction sessions (LLM prompts/responses)
- Architecture plans and endpoints