Tano identifies local documents and attachments using a workspace UUID and short file IDs. A path can change while identity follows the file where it can be determined; Markdown content remains an ordinary file.
Where identity is stored
.tano/config.json holds only portable workspace identity. .tano/identities.json records short file IDs, paths, content hashes, versions, and related identity data. The search index is device application data and can be rebuilt from files and the identity manifest.
A short ID is not a filename. On rename, move, or path exchange, Tano uses filesystem identifiers, previous paths, and content hashes to match identity. When the evidence is ambiguous it does not arbitrarily assign an old ID to another file. A cross-volume move that also changes both name and content may receive a new ID.
Why integrations need this
References and future collaboration cannot rely on paths alone. Check the workspace UUID and current file ID before acting on a file; also check the version or hash from the read before writing. Local MCP requires expectedHash, and file sync checks remote versions.
Move or copy a workspace
Save and close Tano instances using the workspace before moving the whole directory, including .tano/. Reopen it at the new location on the original device. For an unrelated copy, create a new workspace instead of treating two folders with the same UUID as separate collaboration sources.
If .tano/ is deleted, do not generate a new identity over old relations. See Backup and recovery first.