Back to home
Currently shipping · v1.20.2

Orty changelog — every release.

What shipped, what changed, and what's still rough. Beta buyers get every update free — the app checks for updates automatically.

v1.20.2

Aug 27, 2026

Guided Profile Builder — a blank box becomes an 8-step questionnaire. Answers become structured JSON sent to Gemini; empty achievements are never invented. Review, tone-switch, and copy locally.

Added

  • Guided Profile Builder — 8-step flow (specialty, editing style, experience, software, strengths, clients, achievements, extra) with single/multi-select buttons and Other fields, replacing the empty text area.
  • Structured generation via Gemini 3.6 Flash — answers become JSON (specialty, editing_style, experience, software, strengths, clients, achievements, extra) passed directly to the model. Empty achievements are never fabricated.
  • Rewritten prompt generates Professional bio, About Me, and Client-facing description — natural, concise, buzzword-free, client-focused. editing_style and achievements are integrated only when provided.
  • Review step with summary of all answers, inline Edit per row (jumps back preserving answers and progress), and a non-blocking Profile strength 0–100% bar plus an Add achievements warning when needed.
  • Achievements helpers — examples like Edited 100+ videos / Worked with YouTube creators / Improved retention, plus the note You can skip this if you don't have achievements yet. We never invent achievements.
  • Tone switching — Professional, Friendly, Confident, Creative, Premium Agency with inline tiny descs (LinkedIn…, Personal creator brand…, etc.) that regenerates with the same answers.
  • Copy options — Copy, Copy as Markdown, Copy as Plain Text per section and for the whole profile.
  • Local persistence — answers and review state in orty_profile_answers_v1 and last result in orty_generated_profile_v1 (Your last generated profile → View/Regenerate), restored on reopen with no cloud or DB.
  • Improve Existing Profile — paste any profile for clarity, positioning, and professionalism improvements without changing niche, now also via Gemini.
  • Loading states with timed progress messages (Understanding your experience → Identifying your strengths → Writing profile → Polishing final version) and an empty state that shows an example before generation and the last saved profile after.

Changed

  • AI provider separation clarified — Gemini (gemini-3.6-flash) now powers Profile Builder, legacy profile generation, screenshot/video analysis, and core Orty AI; Groq (text_agent.py) powers Co-Pilot only.
  • Builder UI and polish ship in both Free and Pro; generation remains gated by ai_profile_gen (403 pro_required in Free).

Found an issue in this release? Email getorty@gmail.com — you get the founder, not a ticket bot.

v1.20.1

Aug 25, 2026

Security, privacy-preserving licensing & stability. Pro becomes a real, verified tier: activate once with your license key, keep your anonymity, and every Pro feature gets enforced server-side — not just hidden in the UI.

Added

  • Real Pro verification — Batch, Long Form video audits, in-app email sending, AI profile generation, and multiple portfolios now require an activated license key (Settings → Orty Pro). Verification runs through Orty's license server; the app never touches the database directly.
  • Privacy-first activation — activating sends ONLY your license key and an anonymous SHA-256 device hash. No username, email, machine serials, or user content ever leaves your machine.
  • One license = one device, with a 7-day offline grace period. After grace expires without a successful re-check, Pro features fail closed while Free features keep working.
  • Usage quotas lift automatically once Pro verifies — the 5/day rating cap and Co-Pilot cap no longer apply.

Changed

  • Server-side feature gates added to batch rating, video analysis, email sending, profile generation, and portfolio creation (Free is capped at 1 portfolio server-side) — these were previously locked only in the UI.
  • Every launch generates a random localhost request token required on all POST/PUT/DELETE requests, blocking drive-by requests from malicious webpages against the local API (e.g., forced reset or shutdown).
  • Workspace export sanitization — Resend API keys, SMTP passwords, and inline AI chat keys are stripped from exported backup zips.
  • Frontend fail-closed — if tier status can't be read, the UI assumes Free instead of Pro.

Fixed

  • Unified the version number to 1.20.1 across the app, installer, executable metadata, docs, and buyer assets.
  • Fixed a version-compare bug that could silently disable the update badge.

Found an issue in this release? Email getorty@gmail.com — you get the founder, not a ticket bot.

v1.20-beta

Aug 19, 2026

Stability hotfix plus Outreach Diagnostics: a release about trusting the tool — loading states that can't hang forever, errors that always arrive as errors, and a panel that tells you why outreach stalls.

Added

  • Outreach Diagnostics — a “What's Holding You Back?” panel reads your CRM and flags High Score / Low Reply, Ghosting (>3 in 30d), Pitch Fatigue (>10 sends/48h with 0 replies), and Niche Mismatch. Pure Python, zero AI calls, zero cloud — max 3 actionable insights per session.
  • Interactive Walkthrough — first-run step-by-step guide (Rate → Score → Email → Goals) for fresh installs with no leads. Auto-triggers the Demo Creator when no API keys are present. Escape skips it, completion is persisted.

Changed

  • Version renumbered 1.2-beta → 1.20-beta for correct semantic-version ordering (1.20 > 1.12) so existing v1.12 users are correctly prompted to update.

Fixed

  • Co-Pilot counter now shows the correct 5/5 session limit (was hardcoded 3/3); welcome-screen text updates dynamically and resets every 24 hours — not on restart.
  • Infinite loading spinners eliminated — every loading state auto-cancels after an 8-second safety timeout across Rate, Batch, Log, and Self-Healing tabs.
  • Nav bar overflow fixed — the 9-tab strip is now a scroll container so all tabs stay reachable on narrow viewports without Shift+Scroll.
  • Privacy footer corrected to read “Opt-in telemetry”, matching the shipped opt-in telemetry system (previously said “No telemetry”).
  • “Unexpected token '<'” JSON errors eliminated — explicit handlers for HTTP error codes plus a guaranteed-JSON catch-all mean no API route can return an HTML error page to a fetch().json() call.

Found an issue in this release? Email getorty@gmail.com — you get the founder, not a ticket bot.

v1.1 – v1.19

Earlier beta releases built the core: Rate/Batch/Log tabs with Gemini scoring and Groq drafting, the glassmorphic UI pass, portfolio switching, and the Groq model swap (v1.1–v1.12). v1.2-beta brought the workflow-intelligence layer — Personal Insights dashboard, Demo Creator, the 7-stage CRM pipeline with schema migration, workspace export/import, weekly goals, better empty states, and the anti-hallucination email fix. Detailed notes for those builds lived in the installer changelog and are being consolidated here over time.