Logseq

1. Overview

Open-source outliner with distinct file-based and database graph models; preservation and exit depend on the graph type and export selected.

Category: Knowledge & Office.

2. Version and scope assessed

Version assessed: Source commit 05c3ae73e08498e9e20f1b11c3eac5525e5e588d.

4. Principle profile

5. Fundamental requirements

6. Evidence and verification

• E1: Logseq documents graph backup and export to SQLite, ZIP with assets, EDN, and Markdown. (source), confidence high
• E2: The DB exporter has explicit options for UUIDs, timestamps, shallow copies and ontology exclusion. Node/property/class timestamps are conditional; content-reference exports retain referenced UUIDs, while partial-map merging warns against use for full-graph export. Thus export modes are not equivalent preservation packages. This is a source-level distinction, not an executed Markdown/EDN/SQLite migration. (source), confidence medium
• E3: The Node SQLite backup helper calls the SQLite backup API, removes the destination file when synchronous or asynchronous backup fails, and closes the opened source connection in the file-based wrapper. This identifies error cleanup; it does not establish a complete graph backup including asset files, successful restoration, or safety if a caller supplies an existing destination. (source), confidence medium
• E4: Graphs and backups are stored locally on desktop or in a chosen browser folder. (source), confidence high
• E5: Knowledge work is user-directed, but database sync, plugins, and conflict resolution were not tested. (source), confidence low
• sovereignty: A positive Sovereignty finding requires established E1, E2 and E4 in the declared scope. The existing access/exit/custody record for Logseq does not establish all three minimums. The prior positive finding is withdrawn; recorded storage and portability strengths remain facts, not a verified failure. (source), confidence medium
• longevity: E3 is not established for Logseq's existing essential-record scope. The previous Longevity finding is replaced by Not assessed, not a verified failure or a positive placeholder. (source), confidence medium
• locality: The graph and backups can remain local and do not require a hosted account. (source), confidence high
• transparency: The positive control observations above do not yet establish whole-scope E5. The methodology requires that minimum before a positive Transparency and Control finding; the previous positive finding is therefore withdrawn to Not assessed, not changed to Does not meet. (source), confidence medium
• E3: Exporter UUID/timestamp/ontology options and SQLite backup failure cleanup are concretely described in source. No populated graph including assets, properties, references and file-to-DB conversion has been independently reconstructed. A successful database-copy mechanism alone does not establish complete graph semantics or safe existing-destination recovery. (source), confidence medium
• E5: Backup failure cleanup can remove its destination, while the collected evidence does not establish effective caller destination safety, plugin authority, synchronization-conflict control or inspectable correction for DB graph mutations. Export options and source error handling are not whole-workflow oversight. (source), confidence medium

7. Migration and recovery

Native storage format: Version-dependent Markdown/Org file graphs or SQLite-backed database graphs.

Synchronization or runtime model: Graph-model-dependent local storage and synchronization.

No export or inspection command is recorded for this entry yet.

8. Architectural strengths

Established strengths and their scope are recorded in the evidence and assessment findings on this page.

9. Architectural limitations

• Identity/timestamp/ontology preservation depends on the selected exporter and options.
• SQLite backup error cleanup is not proof of complete asset-inclusive recovery.

10. Assessment findings and remaining work

Verification scenario: Specify the file graph versus the database graph, then export and reopen block references, queries, whiteboards, and tasks.

What may invalidate an overconfident claim: Do not generalise file-based graph behaviour to every Logseq storage mode.

Documentation starting point: https://docs.logseq.com/

Evidence update: 23 October 2026

• Resolved: The DB exporter has explicit options for UUIDs, timestamps, shallow copies and ontology exclusion. Node/property/class timestamps are conditional; content-reference exports retain referenced UUIDs, while partial-map merging warns against use for full-graph export. Thus export modes are not equivalent preservation packages. This is a source-level distinction, not an executed Markdown/EDN/SQLite migration. Source.
• Resolved: The Node SQLite backup helper calls the SQLite backup API, removes the destination file when synchronous or asynchronous backup fails, and closes the opened source connection in the file-based wrapper. This identifies error cleanup; it does not establish a complete graph backup including asset files, successful restoration, or safety if a caller supplies an existing destination. Source.
• Open: A populated DB independently queried, complete asset-inclusive graph export/restore, file-to-DB conversion, task/property/ref fidelity, automatic-backup retention and cross-version outcomes. E5 needs caller destination safety, plugin and DB-sync authority, change inspection and recovery/retention boundaries.

Assessment follow-up: These bounded findings are incorporated into the evidence above. Eligibility and principle ratings are deliberately unchanged; their reconciliation and any blank-status fixes are deferred to the next pass. Reproduction details and retained inputs are in evidence/third-pass/batch-6/.

Human-readable criteria: Can a person preserve content, structure, links, metadata, and history in an independently readable form, then restore and edit it without a vendor account?

Rating reconciliation: 24 October 2026

E3 and E5 remain Not Verified for the existing declared scope. Longevity and Transparency and Control are Not assessed pending their required minimums. Sovereignty is also Not assessed because the required access/exit/custody minimums are not all established. This is a targeted evidence decision, not a finding that the product fails. Existing E1, E4 and Locality findings were not re-certified.

Remaining: A populated DB independently queried, complete asset-inclusive graph export/restore, file-to-DB conversion, task/property/ref fidelity, automatic-backup retention and cross-version outcomes. E5 needs caller destination safety, plugin and DB-sync authority, change inspection and recovery/retention boundaries.

Next without installing locally: Acquire a synthetic populated graph DB plus assets and complete export-option manifest; inspect it independently and seek same-version caller, backup-retention, plugin and sync conflict evidence.

The individual decision and provenance are retained in evidence/rating-reconciliation/batch-7/. Overall: Insufficient evidence.

Individually reconciled in rating batch 7 on 24 October 2026. E3/E5 remain Not Verified for explicit missing coverage; dependent unsupported findings corrected. Previously resolved questions are retained, not counted again. E1/E4/Locality were not re-certified by this targeted review. The assessment procedure is described in the methodology.