Nivune v0.2.0-preview.4

Nivune v0.2.0-preview.4

Preview 4 is the first release under the Nivune name. It keeps relationship analytics local while expanding the product from point-in-time lists into a private audience history.

Highlights

• Original Nivune data-iris identity and application icons
• User-confirmed observation dates and audience trend history
• Comparisons between any two snapshots, smart lists, and per-username relationship timelines
• Authenticated passphrase-encrypted portable backups with validated restore
• Stronger archive, ZIP64, compressed-work, paging, and atomic-export boundaries
• Refreshed task-based documentation with a synthetic, reproducible UI tour

Upgrade notes

On first launch, Nivune looks for the former application database at app.insight.local/insight.db. When no Nivune database exists, it copies that history transactionally into app.nivune.local/nivune.db and retains the original file as a recovery copy. It never overwrites an existing Nivune database.

Backup encryption protects portable backup files. The live SQLite database remains unencrypted and depends on operating-system account and disk protection.

Downloads and verification

The release workflow publishes MSI and setup EXE installers for Windows x86-64, a DMG for macOS Apple Silicon, and AppImage, DEB, and RPM packages for Linux x86-64. Verify the selected file against SHA256SUMS before installing.

These preview artifacts are not code-signed or notarized, so operating systems may show an unknown-developer warning. Do not disable system security controls globally.

Known dependency advisory

GitHub reports the moderate glib advisory GHSA-wrw7-89jp-8q8g in Tauri's transitive Linux GTK dependency line. Static source inspection found no Nivune or framework consumer calling the affected VariantStrIter iterator API. The alert remains open because the current Tauri GTK stack cannot yet resolve to the fixed glib major version. This is a bounded prerelease risk assessment, not a claim that the dependency is patched.

Fame boundary

Fame remains privacy-engineering research. Public-profile retrieval, ranking runs, and Fame networking are unavailable; the application offers no direct Instagram access or fallback transport.

Windows and macOS packages are cross-built in GitHub Actions. Manual launch verification is reported separately and should not be inferred from a successful build alone.

insIGht v0.2.0-preview.3

insIGht v0.2.0-preview.3

This unsigned preview improves large-export responsiveness, archive hardening,
report safety, and release reproducibility while preserving the offline product
boundary.

• Load relationship and adjacent-snapshot change results incrementally with
  debounced native search.
• Compute summaries with bounded-memory SQL and stream CSV/JSON reports from a
  transaction-consistent database snapshot.
• Write reports through private temporary files and atomically replace the
  selected destination only after a successful flush and sync.
• Reject oversized or malformed ZIP/ZIP64 metadata, excessive compressed work,
  duplicate canonical paths, unsafe paths, and excessive relationship records.
• Use one pinned Rust workspace, toolchain, and lockfile across the desktop,
  Fame core, and synthetic lab.
• Gate release packaging on the complete build, frontend test, Rust format,
  strict Clippy, locked Rust test, dependency-audit, and synthetic-lab checks.
• Validate the import and SQLite persistence path against a private real export;
  no private fixture or account data is included in this release.

Fame networking, retrieval, and ranking remain unavailable. The app contains no
Instagram login, scraping, credential, PIR, mixnet, telemetry, or fallback
network path.

Artifacts are unsigned. Verify downloads using the attached SHA256SUMS file.

insIGht v0.2.0-preview.2

insIGht v0.2.0-preview.2

This unsigned preview completes the hardened offline workflow on main.

• Select any historical snapshot and compare it with its immediately prior import.
• Export relationship lists or change reports as versioned JSON or CSV.
• Rename local account histories or delete an account and all of its snapshots.
• Import complete synthetic-tested ZIPs and folders through native path mediation.
• Inspect the Fame engineering roadmap without enabling any network retrieval.
• Run the synthetic-only Fame feasibility lab; unmet external gates fail closed as BLOCKED.

Fame retrieval and ranking are not available. The desktop application contains
no direct Instagram client, credential path, PIR transport, mixnet transport,
or fallback network path.

Artifacts are unsigned. Verify downloads using the attached SHA256SUMS file.

insIGht v0.1.1

First public release

insIGht is a local-first desktop application for analyzing official Instagram JSON exports without logging in, scraping, telemetry, or uploading account data.

Included

• Followers, following, mutuals, and non-reciprocal relationship lists
• Searchable account data and local multi-account snapshots
• Change detection between imports
• Duplicate snapshot detection and snapshot deletion
• CSV and JSON exports
• Defensive ZIP and folder import validation

Downloads

• Windows x64: MSI or setup EXE
• macOS Apple Silicon: DMG
• Linux x86_64: AppImage, DEB, or RPM

Important

These artifacts are not code-signed or notarized. Windows and macOS may show an unknown-developer warning. Verify each asset using the SHA-256 digest shown by GitHub before installation.

See CHANGELOG.md for the complete change history.