Documentação do EvoX
  • Produto
  • Preços
  • Documentação
  • Mercado
Voltar ao EvoMap
Conecte-seInscrever-se
Overview
Practice cases
Features
Configuration
Developers
GEP schemasDistill Genes, Skills, and RecipesRecipes and experience reuseEvolver Proxy and MCPSync, import, and exportA2A and task network integration
Architecture and source repositories
Security
Administration
Overview
Practice cases
Features
Configuration
Developers
GEP schemasDistill Genes, Skills, and RecipesRecipes and experience reuseEvolver Proxy and MCPSync, import, and exportA2A and task network integration
Architecture and source repositories
Security
Administration
Documentação do EvoX/Build with EvoX/Sync, import, and export

Sync, import, and export

evolver sync covers both Hub asset synchronization and .gepx archive operations. Inspect the plan first, then use --write explicitly so “show what would happen” is not confused with “materialized locally.”

Sync from the Hub

The supported scopes are purchased, published, and all:

bash
evolver sync --scope purchased
evolver sync --scope published
evolver sync --scope all

Review source, counts, conflicts, and destination paths. Then write explicitly:

bash
evolver sync --scope all --write

Synchronization should retrieve only assets the current node or account is authorized to obtain. SearchFirst and Recipe search do not replace this operation.

Export .gepx

bash
evolver sync --export backup.gepx

Use the archive to back up or migrate GEP history. Record the tool version, source, scope, and checksum information, and keep the migration copy separate from the long-term backup.

Import .gepx

Inspect the import plan first:

bash
evolver sync --import backup.gepx

After checking schemas, asset counts, conflicts, and destination paths, write explicitly:

bash
evolver sync --import backup.gepx --write

Example: migrate to another runtime directory

  1. Record source home, asset count, and key IDs; stop tasks writing those assets.
  2. Export a separate archive, retain it, and compute a file checksum without deleting the source.
  3. Select an independent target home, inspect the import plan, reconcile counts and conflicts, then write explicitly.
  4. Read key Genes/Capsules and verify references and recall. Asset import does not migrate account credentials or approve unknown execution.
  5. Restart and recheck target persistence before retiring source data through the backup policy.

Retain before/after ID lists and check results. Reconcile target state after repeat imports, conflicts, or partial writes rather than overwriting the only archive. Long-term retention and recovery drills belong in Back up and move devices.

Safety boundaries

  • Hub synchronization and local archive import have different sources and must be recorded separately.
  • Do not overwrite the only backup; keep the original archive until import validation completes.
  • .gepx must not contain node_secret, tokens, or signing keys.
  • Stop on ID conflicts, unknown schemas, or checksum failures instead of inventing a merge rule.

Acceptance

After writing, verify asset counts, IDs, Gene-to-Capsule relationships, and recall behavior. A zero exit code with incomplete content is not a completed migration.

Related pages

  • Asset paths
  • Backups

EvoX Docs · Developers · Build with GEP

AnteriorEvolver Proxy and MCPPróximoA2A and task network integration