A record of changes, refinements, and small delights.
57 entries·10 months·last update November 8
·
November 8 Update
Fixed
Trust this device — when you sign in, ticking “Trust this device for 30 days” now works whether you tick it before or after typing the verification code, because the code is now sent only when you click Verify or press Enter.
Common ancestors in DNA Evidence Review — when two people are related through more than one path, “How They’re Connected” now names the right common ancestor for each path, also when one side of the path has more generations than the other, as with an uncle and his nephew; a pair checked before this update shows its paths by number until you click Run Check again.
Research Prompts in a copied POI project — when you duplicate a POI project, the copy now shows its Research Prompts right away instead of only after its next recalculation.
Smaller fixes and polish — when sending feedback fails, when the Security tab in Settings cannot show your trusted devices or revoke one, or when your answer to “Tell us about your research” in a new project cannot be saved, you now see an error message and can try again instead of an empty list, no response or a button that keeps spinning; the feedback form also stops at the longest title and description it accepts.
November 6 Update
New
What each relative’s DNA says — in an endogamous POI project, the table of tested relatives for each placement now has a DNA term column that shows whether each relative’s DNA supports the placement, contradicts it or says little about it; in every POI project, the DNA Fit tooltip now also lists that relative’s shared segments: the total, the part that counts, the longest and the average.
Check the recorded parents with DNA — a new switch in a POI project’s settings ranks the placements as if the person’s recorded parents were unknown, so after Recalculate you see whether the DNA still puts their line first or how many placements score higher, which is a reason to check that link, not proof that it is wrong.
Improved
More reliable DNA from raw DNA files — when Kinprove compares the raw DNA files you uploaded, segments in three stretches of chromosomes 1, 8 and 10, where many people share DNA that does not point to a recent common ancestor, no longer count; DNA imported from MyHeritage or FamilyTreeDNA match lists is not affected.
Half relationships through the wife — in a POI project, placements with a half relationship now also follow another marriage of the wife when that marriage is recorded in your tree or you entered it, not only another marriage of the husband.
Projects recalculated — every project with uploaded raw DNA files and every POI project is being recalculated for this update, so shared DNA totals and the order of your placements may change.
One vocabulary in French, Italian, Spanish and Portuguese — the app, the Chrome extension, emails, the handbook, the privacy pages and earlier news now use one approved name for each genealogy and DNA term in these languages, chosen after checking how established genealogy and DNA sites name it.
Faster public pages — the screenshots on the home, features and handbook pages now arrive in a size that fits your screen, so these pages open faster and no longer jump while the images load.
Clearer search results — the home, features, pricing, news and handbook pages now have titles and descriptions that say what each page offers, in all eight languages.
Fixed
Clear permissions for AI assistants — when you connect an AI assistant, the permission screen now says plainly that it can read, create, change and delete your projects, family trees, DNA kits, DNA segments and hypotheses without Kinprove asking you to confirm each change, and that what it reads goes to the company that provides it.
Smaller fixes and polish — the security note in the sign-in code email now opens the Account section of Settings instead of the home page, public page addresses open directly, an address that does not exist now shows the “Page Not Found” screen, the social sign-in buttons on public pages lead to the right address, a new page now opens at the top instead of part-way down, and long names in the table of tested relatives now wrap instead of being cut off.
November 4 Update
Improved
Dates in the placement ranking — every POI project now checks the birth years a placement implies against the ages at which parents in your own tree usually had children, or against a general pattern when your tree has too few dates: placements inside the usual range lose nothing, and only those that need an unusually young or old parent rank lower.
DNA in endogamous projects — in an endogamous project or POI project, each tested relative’s shared DNA now raises or lowers a placement by how well it fits, compared with the DNA that people in such a community share even without a recorded link, instead of a fixed bonus for every connected relative.
All POI projects recalculated — every POI project is being recalculated with these rules, so the order of your placements may change.
November 4 Update
New
Security tab in Settings — Settings now has a Security tab where you see your trusted devices, named by browser and system, and can log out on every device at once.
Improved
Privacy and Terms — Kinprove now loads its fonts from its own servers, so opening a page no longer sends your IP address to Google; the Privacy Policy says so in every language, and the Terms of Service are now published in English only.
Help and the AI assistant guide — the Help item in the app and the link in the site footer open the Support page, and the guide to connecting an AI assistant is now in every language and explains how to connect with an API key.
Several kits for one person — POI projects show each kit’s testing company, and Kinprove now refuses a second kit from the same company for one person, also when you merge two people.
Fixed
Kit comparisons when you set up a POI project — the participant advice and the close-relative warning now count DNA found by comparing your uploaded kits with each other, not only imported match lists, and each participant’s DNA tag now matches the DNA Evidence Coverage panel.
POI Tree with one person of interest — it now shows the recorded ancestors up to ten generations instead of only the parents, and it keeps the branches you collapsed.
Smaller fixes and polish — Logout everywhere now works for accounts that sign in with Google or Facebook, the password strength meter follows the real rules (at least 12 characters), kit pages say “not measured” for X-DNA that was never compared, and error messages are fully translated.
November 2 Update
Improved
See why a placement is not ranked — instead of a bare “Not ranked”, the “Your lines” card and the “Why this position?” panel now give the reason: you blocked that branch, none of the DNA comparisons can test the placement, or the DNA does not fit it, and then they also name the relative whose shared DNA falls outside the predicted range.
Fixed
Lines that end above the POI are scored correctly — when you build a line by hand to test who the parents of one of the POI’s recorded ancestors were, or of a person you entered above one, the POI is now scored at its own generation below that person, not as if it stood in that person’s place, and X-DNA counts only if it can pass down that line; the few POI projects this affects are being recalculated.
Results from earlier settings are marked — on the Raw DNA tab, a kit whose matches were found with earlier settings now reads “Results from earlier settings”, with a note to press Force Recalculate, instead of showing “Not compared to other kits yet” next to its matches.
The match graph keeps up — the match graph on the DNA Data page now reloads after you upload a kit and again when DNA processing finishes, instead of showing the old matches until you reload the page, and it keeps the zoom and position you chose.
Smaller fixes — the date on a project’s card now shows when its family tree or DNA last changed, not when its settings were last saved; on the Match Data tab, the confirmation for deleting all segments of a person now names the person and says what will be deleted; the hypothesis tree no longer shows an error screen when it reloads after a Recalculate; and if something does go wrong, you now see a plain notice instead of technical details.
October 30 Update
New
Leave theories out of your hypotheses — a new “Theory records” setting in each POI project lets you choose “Leave out”, so hypotheses are built as if the people, couples and parent links you marked “Theory (unconfirmed)” were not in your tree, or keep “Use as premises”, where they are used and carry the “Premise” mark, as before; the change takes effect the next time you press Recalculate, and your assistant in Claude or ChatGPT can change the setting too once you disconnect and reconnect the Kinprove connector.
Fixed
Hypotheses follow your recorded tree more closely — when your tree records that a POI descends from a couple three or more generations back, that couple is now proposed as the POI’s ancestors only at the generation your tree records; a combined hypothesis can no longer contradict what your tree records about the closest common ancestors of two POIs; and a line through a half-sibling is now checked against the dates of the parents it descends from, not those of the other spouse; the POI projects these rules affect are being recalculated.
A failed Recalculate stays visible — when a Recalculate fails and the page then cannot reload the project, the POI page and the hypothesis tree now keep showing the notice with the reason and the Recalculate button, instead of the results from before the run or a loading error.
October 29 Update
New
Mark what is still a theory — when you edit a person or a couple in your family tree, you can now tick “Theory (unconfirmed)” on the person, on the couple, or on the person’s descent from their father or mother; every hypothesis whose line runs through a marked record then carries the “Premise” mark, just like a line through a person you entered by hand, and the mark appears at once, without recalculating.
For Claude and ChatGPT users — your assistant can now find people in a hypothesis tree by name or by their id in your family tree without loading the whole tree, rule out a couple as common ancestors even before any hypothesis uses them, open large hypothesis trees faster, and compare two placements on the relatives both of them connect; disconnect and reconnect the Kinprove connector, then start a new chat, to get the new tools.
Improved
Recalculate runs in the background — Recalculate now responds at once and shows “Recalculating…”; you can keep working, the results appear by themselves when it finishes, and pressing it again in another tab no longer starts a second run.
Fixed
Recorded dates now rule out impossible placements — a placement is no longer proposed when the birth years of the POI’s recorded parents make it impossible, and a combined hypothesis for several POIs is set aside when the recorded line between two of them would leave too few or too many years between the generations; POI projects are being recalculated with this check.
Each POI’s own share in a combined hypothesis — when a hypothesis places several POIs at once, the POI card, the project list and the hypothesis list in the tree now show each POI’s own share, for example “17.2% · 13.0% · 22.4%”, instead of one combined score that looked like a probability; and the warning that a leading placement is ahead only because it connects more relatives now also appears when the leader scores lower on the relatives both placements connect.
October 28 Update
New
A birth year for each POI — a POI project with several POIs now takes a birth year for each one, and adding another POI keeps the year you already entered.
Improved
Save now, recalculate once — saving a change to a POI project, such as its POIs, its relatives or its settings, now takes seconds instead of rebuilding every hypothesis while you wait; the project then shows “Configuration changed — press Recalculate”, and one press of Recalculate brings the results up to date, keeping the lines you built by hand and scoring them again.
Hypotheses are built and scored much faster — on a large family tree, Recalculate now finishes in a few minutes instead of more than half an hour.
Fixed
Genotek DNA files are recognised — Kinprove now tells a raw DNA file from Genotek apart from a 23andMe file, which it closely resembles, and compares the two with the stricter settings meant for two different testing companies, so fewer short false segments remain; Genotek files uploaded earlier are being checked again, so some shared cM totals with those kits may go down slightly.
Smaller fixes and polish — the DNA Evidence Review now shows when a check is already running and says why a new check could not start; a page left open while Kinprove was being updated now reloads itself once instead of showing an error; a large POI project can now be copied; and many texts in German, Ukrainian, Russian, French, Italian, Spanish and Portuguese read more naturally.
October 26 Update
New
See why a hypothesis ranks where it does — a new “Why this position?” panel compares a placement with the top-scored one, part by part, and names the relatives that only one of the two connects; placements whose expected DNA is identical are shown together as one group; “Your lines” on the Top Hypotheses card shows what the DNA says about each line you built by hand; the coverage panel shows, for each relative and POI, what DNA evidence there is and why a pair has none; and for a relative connected through several lines, the expected DNA now explains how those lines add up.
A code by email when you sign in — signing in with your email and password now asks for a 6-digit code we send to your inbox; tick “Trust this device for 30 days” to skip it on your own computer, and signing in with Google or Facebook works as before.
Photos in your family tree — add a photo to each person in their edit dialog and see it on their card in the tree.
A map of your DNA matches — a new Graph tab in the DNA library shows everyone in the project who has a DNA match and how they are connected, with a slider for the smallest shared cM to show.
For Claude and ChatGPT users — your assistant can now find duplicate people, check your tree for impossible dates, read one person’s full details, show every segment behind a pair, find whose imported segments overlap a region of a chromosome, import a MyHeritage or FamilyTreeDNA segment file for people already in your tree, and explain why a couple was not proposed; disconnect and reconnect the Kinprove connector, then start a new chat, to get the new tools.
Improved
DNA matching is noticeably more accurate — every segment Kinprove finds between two raw DNA files now passes an extra marker-by-marker check, and a segment that does not hold up is dropped, so fewer false short matches remain while close relatives keep nearly all their shared DNA; an X-DNA segment now counts only from 10 cM, in every project; files from testing companies that use the newer genome reference (GRCh38) are now converted and compared like any other, and a file whose genome reference cannot be confirmed is named in your DNA library and left out instead of producing wrong matches; a pair of relatives that has not been compared yet now reads as unknown instead of as sharing no DNA; and the DNA in every project with raw DNA files is being recalculated with these changes, so match counts and shared cM may change as the recalculation finishes.
Realistic expectations for your relatives — how much DNA two relatives are expected to share now comes from simulations of DNA passed down through many generations of families, using the same minimum segment size as your matches, so the expected ranges for distant cousins, including the real chance that they share no DNA at all, are more realistic in your DNA Evidence Review and in your hypotheses; your Evidence Reviews are being run again for you; and a POI project scored as endogamous now counts only segments of at least 12 cM, or your own stricter cutoff, instead of widening the expected range, with a new preset that sets this up in one click.
Hypotheses find more of the right ancestors — Kinprove now considers every ancestor couple of even a single tested relative, not only couples at least half of your relatives share, reaches one generation further back, follows the ancestor families recorded in your tree further up, and also builds on families with only one known parent, on parents you entered yourself (marked “Premise”), and on places directly below a tested relative, such as an adoptee whose grandmother tested; after the DNA recalculation each POI project offers to regenerate its hypotheses, which adds these new candidates but replaces the lines you built by hand, while editing your POIs or the list of relatives no longer erases them.
Kinprove Ltd — Kinprove is now run by Kinprove Ltd, an Irish company; the Terms of Service and the Privacy Policy are now available in every Kinprove language and say plainly that your raw DNA files are never shared with third parties; the company details now sit in the page footer; and the Support page has been rebuilt with a contact card and answers to common questions in every language.
A smoother DNA library — processing progress now appears the moment an upload finishes; a kit with an empty or unusable file carries a badge; FamilyTreeDNA now tells you when a new kit’s matches are still being prepared, and you are warned before leaving an unfinished import; an imported match filed under the wrong person can be moved with “Wrong match — different person”; and the browser extension now brings in MyHeritage triangulated segments and tells you when a capture fails.
Fixed
Fairer scores for your hypotheses — each relative is now weighed through their own closest shared ancestor and their own full or half relationship, the same line of descent is no longer counted twice, triangulation counts only groups that include the person you are researching, the birth year you entered for a POI is used, an undated couple is dated from its descendants, placements with equal scores share a rank, a relative you marked as sharing no DNA no longer adds a coverage bonus, and each POI project can set its own stricter minimum segment size; your POI projects are being rescored for you.
Smaller fixes and polish — removing or adding a parent for one child no longer changes that child’s brothers and sisters; a GEDCOM file with no people in it can no longer empty your tree; names Kinprove gives to unknown ancestors, and relationship names, now appear in your language; pages fit better on phones and tablets, and the Feedback button has moved into the sidebar; many German and Russian texts read more naturally; and sign-in and account security have been tightened in several places.
August 30 Update
Improved
A way forward when your file has no X — a kit marked as carrying no X-DNA now points you straight to the segment import, where you can add a match list that does carry it.
Fixed
X-DNA from an imported match list now reaches your results — when a match list you brought in from FamilyTreeDNA carried X-chromosome sharing, some of it was quietly set aside on the way in and never counted; it now comes through.
A kit with no X-DNA no longer hides what your matches show — some raw DNA files hold no X data at all, and Kinprove used to read that as “nothing shared” on the X chromosome even when an imported match list said otherwise; every chromosome is now weighed on its own, so imported X sharing is used wherever the raw file has none.
August 30 Update
New
Kinprove now looks for DNA kits filed under the wrong person — your kits are weighed against one another, so you are told when two files contain the same genome, which means either the same sample filed under two people or identical twins, and when one person’s two kits share very different amounts of DNA with the same relative, so they cannot both be that person; the findings appear on your DNA Evidence Review, the card on your Projects list shows how many were found, a kit that could not be fully checked is named with the reason in plain words, and a kit now carries a note when no X-DNA could be read from its file or when much of the file is blank.
Add a brother or a sister straight from the tree — a sibling can now be added from the person themselves, with gender and last name already filled in, instead of navigating to the parent and adding a child there; and the menu of actions on a person opens on whichever side has room, so every option stays within reach.
Improved
A safer re-import — bringing a GEDCOM file into a project that already holds a tree now asks you to confirm first, and lists exactly what would go: how many people and families would be deleted, and how many DNA kits unlinked; a MyHeritage match list captured by the browser extension is no longer refused over blank rows; and a part-finished upload now tells you how many files were saved and keeps the ones still needing your attention on screen.
Matches come back sooner, and read truer — comparing two kits with one another, and working out three-way triangulations, both take noticeably less time now, so a big project finishes its analysis sooner while finding exactly the same shared segments; the shared-DNA totals taken from imported match lists now count only the 22 numbered chromosomes; and the probabilities on your ancestor hypotheses read the same wherever you meet them.
Fixed
A pedigree that lines up — when one person has to sit in two irreconcilable places in one pedigree, such as a man married to both an aunt and her niece, or two cousins who marry, the second place now gets its own card with a small ×2 badge, so generations and rows of siblings line up instead of being pulled out of position.
Smaller fixes and polish — dates that came in from a GEDCOM file with a qualifier now read as “c. 1850”, “bef. 1900” or as a range instead of a bare year, and a year you entered on its own stays a year; one DNA file Kinprove cannot identify no longer holds up the search across the rest of your project, which goes ahead with the files it recognises while your DNA library names the one it left out; and reliability work continues across the DNA browser, family tree, and research projects.
August 18 Update
New
Merge duplicate people without leaving Kinprove — when the same person appears twice, you can now merge the two records from the DNA library itself instead of going through Claude or ChatGPT; the dialogue tells you how many DNA kits will move across, asks you to confirm which record is kept, and warns you when a family link could not be carried over.
Marriage and divorce dates on the line between a couple — click the line joining two people to set or change their marriage and divorce dates; a couple with a divorce date is now drawn with a dashed line.
Improved
Kinprove now tells you when a DNA analysis finds nothing — a kit that produced no matches, or a project whose kits can never be compared with one another, is now reported to you with the reason instead of leaving you looking at an empty screen; the DNA library shows a notice at the top of the list when something needs putting right, such as several kits assigned to the same person or a kit owner who is not connected to your tree, and every kit now shows how many of its comparisons have been worked out and how many matches came back.
Enter dates the way you write them — the date field now follows your own language and reads back what it understood — “4 March 1920”, “March 1920”, “1920” — and flags anything it could not make sense of instead of quietly ignoring it; editing someone’s details no longer shortens a full birth date to just the year, and years now show correctly on the cards in your tree.
Fixed
A DNA file Kinprove cannot recognise is now questioned, not turned away — instead of being rejected and lost, the file stays on the upload screen so you can tell Kinprove it really is raw DNA data; before anything is saved, the review step names exactly what will be stored; and files are identified far more accurately, so a document that is not DNA data can no longer end up filed under a testing company it never came from.
When a person cannot be deleted, Kinprove explains why — you now see what is holding them — children, DNA data, research projects — with counts and a way to reach each place where you can clear it, instead of a delete that quietly does nothing.
Smaller fixes and polish — matches that have already been worked out are no longer hidden behind the wrong tab, the dashboard no longer counts the same match twice, someone connected to your tree by marriage is now recognised as part of it, and reliability and security work continues across the DNA browser, family tree, and research projects.
August 14 Update
New
Choose the person your tree starts with — the first time you open each of your projects, Kinprove now asks you to pick a start person, so if you see that question instead of your tree, nothing is missing: choose someone once and the tree opens centred on them from then on, and you can change your choice later in the project's settings.
Bring in FamilyTreeDNA matches without a file — sign in to FamilyTreeDNA from Kinprove, search your matches in a list that puts the strongest first, tick the ones you want, and link them all to people in your tree in one go.
See which triangulations are still missing — a research project now shows how many groups of three have been triangulated and, when a group's triangulation cannot be worked out from the data you have, names the MyHeritage matches worth exporting again, with a copy-ready list and a way to dismiss the ones you do not need.
Improved
A rebuilt family tree — every tree now opens as a pedigree, with each generation in its own tidy column, tall portrait cards, and each marriage of someone who married more than once drawn as its own couple, so a Kinprove tab you left open from before this update may look blank or broken until you reload the page once.
Changing your password now signs out your other devices — for your safety, changing your password signs you out everywhere except the browser you changed it in, while resetting a forgotten password signs you out on every device including that one, and either way the keys that connect Kinprove to Claude or ChatGPT are cancelled, so if you use an AI assistant, create a fresh key under Settings > API Keys and reconnect your assistant with it.
Fixed
Your whole GEDCOM file comes through — opening your family tree straight after a GEDCOM upload no longer freezes the page, everyone from the file now appears in your tree, and people brought in from a family file no longer receive invented surnames, with any wrong ones already in your tree corrected for you.
Smaller fixes and polish — the full analysis now covers up to 60 DNA kits in a project instead of 50; Kinprove can merge duplicate people for you through Claude or ChatGPT; your first DNA upload is guided step by step; large trees appear faster; the wording is clearer when an import or a FamilyTreeDNA sign-in goes wrong; and reliability work continues across the DNA browser, family tree, and research projects.
July 14 Update
Improved
See what your imported matches added — the DNA segment-import history now shows how many triangulations were found among the matches you brought in from another testing company.
Faster on large trees and research projects — Kinprove does far less work behind the scenes when building big family trees and working out ancestor hypotheses, so they load and update more quickly.
Fixed
Smaller fixes and polish — a clearer message when an uploaded file isn’t raw DNA data, plus reliability and security work across the DNA browser, family tree, and research projects.
July 7 Update
New
Triangulations from your imported matches — Kinprove now finds triangulations among the DNA matches you bring in from MyHeritage, not just from raw DNA files. They appear on the chromosome browser and are counted for each person, so you can see who shares the same stretch of DNA even when you only have match data. Turned on automatically.
Sign in and privacy — link or unlink your Google and Facebook accounts from settings, and a cookie notice now appears on your first visit.
Improved
A tidier picture across testing companies — when someone has tested with more than one company, Kinprove now blends their shared segments together sensibly, favours the more detailed source, shows which company each part came from, and keeps everything aligned when you move or merge matches.
A smoother DNA library — select every kit with a single tick, see a person’s full name and birth year when a name is shortened, spot the active kit at a glance, and read relationships in a compact, consistent form.
Fixed
Large family trees import in the background — bringing in a big GEDCOM file no longer leaves you on a frozen page; the import runs in the background with a progress indicator, so even very large trees come in dependably.
Smaller fixes and polish — more dependable DNA analysis on large collections, a smoother FamilyTreeDNA sign-in, more reliable email delivery, plus security hardening and reliability work across the DNA browser, family tree, and research projects.
June 24 Update
New
Record divorces and year-only dates — you can now note a divorce date on a marriage, and enter a marriage date as just a year when that is all you know.
Improved
A clearer DNA library — when someone has tested with more than one company, their kits now sit in one tidy card with a sub-card for each testing company that shows how many triangulations it takes part in; the Match Data tab breaks a person’s matches down by company and lets you remove the matches from any one of them; and the Raw DNA list is simpler to scan.
Know when a hypothesis is out of date — in research projects, the notice on your ancestor hypotheses now tells you whether your family tree or your DNA changed, and it updates whenever you delete or move imported matches.
Fixed
Steadier totals across a person’s kits — when someone has tested with several companies, Kinprove no longer counts the DNA they share more than once, so match totals and relationships stay accurate.
Raw DNA results always appear — the Raw DNA view no longer gets stuck showing “Calculating…”, so your shared-segment results come through.
Smaller fixes and polish — nightly maintenance now runs dependably to keep your data tidy, plus security hardening and reliability work across the DNA browser, family tree, and research projects.
June 17 Update
New
Sort by birth year — siblings and people in your family tree can now be ordered by year of birth, so the chronological view comes naturally.
Improved
DNA analysis starts on its own — when you assign a DNA kit or bring in shared segments, Kinprove begins the analysis automatically, so your results are ready without you having to start it by hand.
Cleaner family-tree toolbar — the relationship-types legend and the biological-only switch are gone, leaving a calmer toolbar with just the actions you reach for.
Tidier public site — the Home page and the User Handbook now appear in the top menu of every public page, the navigation in longer languages no longer overlaps the hero logo, and the Kinprove logo on sign-in always takes you back home.
Real screenshots on the features page — the Claude Desktop conversation and the browser-extension side panel now show actual pictures instead of placeholders.
Fixed
Smaller fixes and polish — the family-tree expand button now reliably opens the right branch, connecting Kinprove to Claude or ChatGPT follows through cleanly, plus security hardening and reliability work across the DNA browser, family tree, and research projects.
June 14 Update
Improved
More accurate DNA matches — Kinprove reads the fully-shared parts of your DNA more precisely and merges shared segments more cleanly, so close relatives and shared-segment details come out more accurate.
Fixed
Your whole DNA file is read on upload — Kinprove no longer skips the first line of certain raw DNA files, so all of your data comes through.
Steadier ancestor hypotheses — in research projects, candidate ancestors are scored more consistently, so your hypothesis rankings hold steady from one run to the next.
Smaller fixes and polish — selecting DNA matches behaves correctly, plan limits explain themselves more clearly, plus security hardening and reliability work across the DNA browser, family tree, and research projects.
June 13 Update
Improved
Faster research projects — duplicating a research project and generating hypotheses is noticeably quicker, even with many people involved.
Smoother sign-in when connecting an AI assistant — linking Kinprove to Claude or ChatGPT now takes you straight to sign in when you are not already logged in, then continues the connection automatically.
Fixed
Steadier DNA analysis on large collections — analysing a large set of DNA matches now runs dependably through to the finish instead of stalling or starting over, so your results arrive reliably.
June 12 Update
Improved
Behind the scenes — security hardening and reliability improvements across the app.
Fixed
More accurate DNA totals — shared centimorgans and segment counts no longer double-count overlapping segments, so your match statistics are exact.
Steadier displays — pages showing people, matches, and triangulations stay stable even when a record is missing some details.
June 9 Update
New
Kinprove now speaks French, Italian, Spanish, and Portuguese — alongside English, German, Russian, and Ukrainian. The whole app, the browser extension, and the user handbook are translated, and each language now has its own web address so you can share a page in the language you read it in.
Jump straight to a match's source — DNA matches you bring in with the browser extension now link back to their MyHeritage or FamilyTreeDNA page in one click.
Improved
More accurate DNA hypotheses — relationship scoring is more precise across the board, especially for families from communities where the same families intermarried over generations, and for people who share more than one line of descent.
Full sibling vs. half sibling, told apart — Kinprove now reads the deeper “fully-shared” parts of your DNA, so it can tell a full sibling from a half sibling far more confidently.
Using Kinprove with an AI assistant? — if you've connected Kinprove to one, please disconnect and reconnect it, then start a fresh conversation — that's all it takes for it to pick up these updates.
Fixed
Adoptive, step, and foster links shown clearly — these relationships now appear distinctly in your tree, with a note that they don't carry shared DNA, so a relative who connects through a non-biological link no longer looks like a missing result.
Smaller fixes and polish — a steady stream of them across the DNA browser, the family tree, research projects, and the language switcher.
June 1 Update
New
Move or merge a person’s DNA matches — right from the Match Data tab. Use the new pencil button on any match card to merge someone who ended up listed twice, or to relink a list of matches that actually belongs to a different person. Kinprove moves the shared segments across and tidies up any duplicates so nothing is double-counted.
Improved
Kinprove is faster — pages load and respond noticeably quicker across the board, especially with large family trees and long DNA match lists.
Sharper relationships in research projects — shared-ancestor and relationship detection now works through proposed ancestors, and those ancestors show by name.
More of Kinprove in your language — the user handbook now reads in German, Russian, and Ukrainian too, alongside more pages and smaller polish across the app.
Fixed
Search handles tricky names — searching now works smoothly even when the text includes punctuation, dates, or symbols, like pasting a name together with birth and death years.
More dependable DNA recalculation — recalculating a project that mixes uploaded and imported matches now runs smoothly.
May 30 Update
Improved
More accurate matches for endogamous ancestry — if your family comes from a community where the same families intermarried over generations, Kinprove now weighs those DNA matches more accurately.
More of Kinprove in your language — the DNA and upload screens, research-project tools, on-screen notifications, dates, and even your sign-in verification email now appear in your chosen language, on top of the menus and buttons.
Fixed
Grab DNA segments straight from MyHeritage and FamilyTreeDNA — the Kinprove browser extension now reliably brings your DNA matches' shared segments from MyHeritage and FamilyTreeDNA into your project, so you don't have to copy them in by hand.
Smaller fixes and polish — a range of smaller improvements across the family tree, the language switcher, and research projects.
May 29 Update
New
Kinprove now speaks German, Russian, and Ukrainian — as well as English. Choose your language from the language menu in the top corner, and the whole app — menus, buttons, messages, and the Chrome extension — switches over right away. Kinprove remembers your choice for next time.
May 29 Update
Improved
More dependable DNA analysis — DNA triangulation is now more thorough and consistent: duplicate segments are cleaned up automatically, fewer analyses get stuck, and a banner shows you when your DNA is still being processed — so your shared-segment results are cleaner and more accurate.
Fixed
Smoother FamilyTreeDNA connection — importing directly from your FamilyTreeDNA account now works smoothly with their updated sign-in.
Smaller fixes — several small fixes across the DNA browser, the family tree, and the Chrome extension.
May 28 Update
New
Best DNA kit per person — Kinprove now shows the best DNA kit for each person, lets you choose another one, and updates that choice when kits are added or removed.
May 27 Update
New
See DNA at a glance on your family tree — people who have DNA now show a small marker in the corner of their box, and you can jump straight to their raw DNA or their DNA matches from the menu that appears when you point at them.
Improved
Faster sign-in — opening the sign-in page now takes you straight to a clean sign-in form instead of the full homepage, so you can get back in right away.
Clearer Settings labels — the Settings section showing whether you've linked Google or Facebook is now labeled "Sign-in method(s)", and it's hidden entirely if you only sign in with email and password.
May 26 Update
Improved
One upload for all your DNA — adding DNA used to mean picking between two separate tools: one for your raw DNA file and another for your list of DNA matches. Now there's a single place to add DNA — just choose your files and Kinprove works out what each one is for you, whether it's a raw DNA download from your testing company or a list of matches. You can even add both kinds at the same time.
May 26 Update
Improved
Clearer messages for unusual DNA files — Kinprove now checks your raw DNA file as soon as you upload it. If the format isn't supported (for example, Genotek data re-downloaded through FamilyTreeDNA), you'll see a clear message right away pointing you to the right download to use.
Friendlier processing errors — error messages on the DNA page are now written in plain language, and when the same issue affects several DNA matches it's shown once with a count instead of repeating.
May 25 Update
Improved
Cleaner Settings page — when you open Settings, the menu on the left now shows just your settings options with a clear "Back to app" link to jump back to your tree, instead of mixing in the main menu. The settings content is also centered on the page to match the rest of Kinprove.
May 25 Update
Fixed
Family tree files with non-English names now import correctly — if your GEDCOM file uses Cyrillic, accented, or other non-English letters in names and places, importing it could stop with an error. Kinprove now reads these files in their original form and imports them correctly, whatever language your records are written in.
May 25 Update
New
Connect Kinprove to Claude and ChatGPT — from the new Connected Apps page you can link your favorite AI assistant to your Kinprove account. Once connected, ask it to search your tree, look up matches, or summarize evidence right inside Claude or ChatGPT. You choose what the assistant can see, and you can disconnect any time.
Safe deletes when AI helps — if your AI assistant tries to delete a person, family, project, or DNA kit, Kinprove asks you to confirm before it happens.
Improved
Chrome extension on the DNA Library page — if you import DNA from MyHeritage or FamilyTreeDNA, the page now points you to the Kinprove Chrome extension as a quicker alternative to manually uploading segments.
Fixed
POI tree switcher — the switcher for moving between trees on a POI detail page is back where you'd expect it.
Better ancestor depth on the mother's line — POI relationship paths that go through a mother now show the correct number of generations.
April 23 Update
Improved
Clear message when your session expires — if your session quietly times out while you're working, a friendly "Session expired" notice now appears on the login screen so you know what happened, and you land right back where you were after signing in.
Fixed
Deep-links survive expired sessions — opening a shared Kinprove link (for example a POI project someone sent you) used to drop you on the home page after signing in, losing the link entirely. Now signing in takes you straight to the page you were trying to open — works the same way for email/password, Google, and Facebook login.
April 23 Update
Fixed
Chrome Extension — every Compare click captures fresh data — previously, switching DNA matches on the MyHeritage Chromosome Browser and clicking Compare again would leave the side panel showing the old match's segments. A page reload was the only fix. Now each Compare click is captured immediately with no reload needed.
April 22 Update
Fixed
Smoother first-time experience — after signing up and adding their first person to the tree, some new users were seeing an error message on the Dashboard instead of their project. That's fixed — your project now opens cleanly.
Both parents show up when you edit a person — if you added a father and a mother one at a time, only the first one was appearing as selected when you reopened the edit dialog. Now both parents show up correctly, and existing people with this mismatch are healed automatically on next view.
April 17 Update
New
FamilyTreeDNA segment capture — the Kinprove Chrome extension now grabs DNA segments straight from the FamilyTreeDNA chromosome browser. Select matches, click Compare Relationship, and the side panel fills with their segments ready to assign — the same flow you already use on MyHeritage.
X-chromosome segments from FTDNA — the X-chromosome segments shown in FamilyTreeDNA's "Detailed Segment Data" tab are captured too, including for matches whose raw data doesn't contain X.
Improved
Works across MyHeritage and FTDNA — the side panel adapts to whichever DNA site you're browsing, with the right empty-state hints and "View on..." link.
April 14 Update
New
Inline side panel for DNA assignment — clicking a match on MyHeritage now opens the Kinprove side panel right next to it, so you can read segments and assign them to individuals in your tree without ever leaving the DNA site.
Work through matches in a row — the side panel collects multiple matches as you open them, so you can review and assign a whole cluster in one session instead of going back and forth.
Improved
Current project always visible — the active Kinprove project is shown as a chip at the top of the panel, and the panel tells you clearly when your session has expired and needs a fresh login.
April 11 Update
New
Endogamy mode per POI project — endogamy can now be enabled on individual people-of-interest projects instead of the entire workspace, so you can mix endogamous and non-endogamous branches in the same Kinprove account.
Improved
Cleaner ancestor trees — ancestors above anchor couples are automatically pruned, so hypothesis trees stay focused on the generations where DNA evidence actually lands.
Fixed
Consanguinity above anchor couples — when two ancestors of the anchor couple share a common line, the hypothesis engine now credits multipath scenarios instead of treating them as impossible.
In-laws no longer look disconnected — DNA Evidence Review distinguishes in-law connections from truly floating participants, so the summary reflects who can and can't be scored.
Direct-ancestor participants — a participant who sits directly on the path from POI to ancestor is now handled correctly in ghost hypothesis scoring.
April 5 Update
New
Outlier flag — segments that fall outside the expected cM range for a relationship now carry a visible outlier flag and can be filtered in or out of scoring.
Over-1500 cM warning — if a participant shares more than 1500 cM with the person of interest (suggesting a parent–child or sibling relationship), Kinprove now warns you so the profile isn't treated as a distant cousin by mistake.
AI assistant greatly expanded — the Claude Desktop integration gained 23 new tools covering tree editing, DNA segments, X-DNA, project management, and hypothesis validation, so you can ask for multi-step research help in natural language.
Improved
Richer evidence matrix — the evidence matrix now shows probabilities, segment-level details in tooltips, and the names of the participants behind each triangulation cell — so you can see at a glance why a hypothesis is supported.
Smarter segment coverage — coverage analysis now includes kit-by-kit pairs, segments imported from matches, and a search fallback for participants with partial DNA data.
March 31 Update
New
Free plan with a starter project — new accounts get a Kinprove project created automatically, so you can upload a DNA kit and explore matches without any setup.
First-time onboarding wizard — after signup, a short wizard asks which uploaded DNA kit belongs to you, so relationships can be displayed from your perspective across the app.
Import history — every DNA match import from MyHeritage and FamilyTreeDNA is now recorded with a timestamp, so you can see when each kit was last refreshed.
Improved
Friendly plan-limit messages — if an action would exceed your plan (more projects, more kits, more people of interest), an inline banner explains the limit and links to an upgrade.
March 30 Update
Improved
DNA section reframed around match data — the section is now called "DNA Data" and leads with the matches you research, positioning raw DNA uploads as an enhancement for triangulation. This reflects how most users actually work: raw DNA only for known relatives, match data for everyone else.
New tab layout — the DNA Library defaults to the Match Data tab, with Raw DNA as the secondary tab, renamed from the previous Kits and Segments labels.
Dashboard counts people, not kits — the DNA card on the dashboard now shows the total number of people you have DNA data for, not just raw kits uploaded.
Consistent naming everywhere — sidebar, navigation, empty states, tooltips and handbook copy were all updated in one sweep so the new terminology lines up across the app.
March 19 Update
New
Introducing the Kinprove Chrome extension — install it once from the Chrome Web Store and capture DNA matches straight from MyHeritage and FamilyTreeDNA. No CSV export, no copy-paste — segments flow into your Kinprove project as you browse.
FamilyTreeDNA direct search & import — type a name, pick up to seven matches, and Kinprove pulls their shared segments in seconds over FamilyTreeDNA's native API.
Improved
Sign in with your DNA-site session — the extension rides on your existing MyHeritage or FamilyTreeDNA session, so you no longer share any DNA-site passwords with Kinprove.
Polished FamilyTreeDNA 2FA — the one-time code entry is now split into six cells with auto-advance, a countdown badge, and a clear warning when the code is about to expire.
MyHeritage localisation — the import works on country-specific MyHeritage domains (UK, DE, FR, and more), not just myheritage.com.
March 5 Update
New
X-chromosome scoring — hypothesis probabilities now factor in X-DNA inheritance paths, giving more weight to branches with X-chromosome evidence.
Improved
Improved composite hypotheses — combined hypotheses are now built as proper hierarchical trees, duplicates are removed automatically, and composites no longer leak into single-POI views.
Group ancestors moved — the "Group by ancestor" toggle is now inside the Hypotheses dropdown for a cleaner toolbar.
Fixed
Better tree layout — edge routing for couples with multiple spouses is more accurate, and dragging nodes no longer disconnects edges.
Segment coverage enhancements — coverage analysis now includes all participants (not just kit owners), and bidirectional imported segments are counted correctly.
Social login fix — Google and Apple sign-in no longer fail intermittently.
February 25 Update
New
AI research assistant (MCP) — connect Kinprove to Claude Desktop and ask questions about your projects, hypotheses, DNA evidence, and family tree in natural language. Generate API keys in Settings > API Keys to get started.
Improved
Settings sidebar navigation — the Settings page now uses sidebar navigation instead of horizontal tabs, consistent with the rest of the app. Each section has its own URL, so you can bookmark or share direct links.
Clickable logos — the Kinprove logo on public pages (Handbook, Pricing, What's New) now links back to the login page.
Smoother landing page — the hero section adapts to content height, and scrolling feels less rigid.
Fixed
Steadier tree interactions — hover effects in the hypothesis tree no longer trigger while panning or dragging nodes.
February 22 Update
Improved
Research prompts on project page — after scoring, participant groups that share triangulated DNA but no known common ancestor now appear directly on the POI project detail page, not just inside the hypothesis tree.
Disconnected participants — the DNA Evidence Review summary card now shows how many participants aren't connected to the family tree, explaining why some pairs remain unscored.
February 20 Update
New
Branch grouping — hypothesis branches that share the same ancestor couple are now grouped together, making it easier to compare related scenarios.
X-match animation — hovering an X-chromosome match endpoint in the hypothesis tree highlights the X-inheritance path with a pulsing animation.
Improved
Gender watermarks — tree nodes now display a large faint gender symbol as a watermark, replacing the previous small badge.
Polished tree interactions — action bubble tooltips are now styled and appear instantly, with a refined shadow and accent border.
Smarter scoring — triangulation evidence now uses log-scale bonuses to better differentiate well-supported hypotheses.
February 18 Update
New
Block ancestor couples — mark specific ancestor couples as blocked to exclude them from hypothesis generation. Blocked couples show an orange hatch pattern in the tree.
X-Match column — the DNA matches table now shows X chromosome shared DNA in its own column.
Improved
Cleaner DNA table — columns renamed for brevity (Shared, Largest), numeric values right-aligned for easier comparison.
Improved readability — text across the app increased from 12px to 14px.
Fixed
Low probability display — very unlikely hypotheses now show "< 0.1%" instead of "0.0%".
February 17 Update
New
Duplicate POI project — create an exact copy of a POI project, including all settings, participants, and hypotheses.
Ctrl+drag subtrees — hold Ctrl while dragging a node to move the entire subtree together in the hypothesis tree.
Swap spouses — new action in the hypothesis tree to quickly swap the husband and wife in a couple.
Tree positions saved — your node positions in the hypothesis tree are remembered across page reloads, per view mode.
Improved
Redesigned dashboard — project cards now show key stats at a glance: hypothesis count, scoring progress, and DNA coverage.
Improved focus mode — focused nodes now glow, action bubbles on dimmed nodes are hidden, and clicking any background area clears focus.
Fixed
Better participant handling — match-based participants now correctly appear in hypothesis generation and the tree; selecting a person as POI automatically removes them from participants.
February 16 Update
Improved
Redesigned confirmations — action confirmations use a distinctive botanical-styled dialog.
Fixed
More accurate scoring — hypothesis probabilities now correctly include triangulation evidence for all ancestor couples.
Reliable evidence indicators — the triangulation flag only appears when backed by raw DNA data, and segment status displays correctly for all participants.
Cleaner triangulation data — unreliable pseudo-triangulations from segment file imports have been removed, keeping only verified results.
February 16 Update
New
Telegram notifications — new features and improvements are now posted automatically to our Telegram channel, so you can follow along without checking the app.
February 16 Update
Fixed
Focused hypothesis generation — only your selected participants now drive hypothesis generation and scoring, preventing phantom results from unselected DNA matches.
Better evidence matrix — segment-based POIs and multi-branch trees now display evidence correctly.
Cleaner project view — POIs no longer appear in the participant list, and DNA type tags (Raw DNA vs Segment) are shown accurately.
February 16 Update
Improved
Smarter tree filtering — when you filter hypotheses, branches with no matching scenarios and disconnected ancestors are now hidden automatically.
Larger GEDCOM uploads — you can now upload GEDCOM files up to 50 MB (previously 10 MB).
Fixed
Better coverage recommendations — segment coverage suggestions now correctly reflect all your imported files.
February 15 Update
New
User Handbook — the new handbook, opened from the sidebar Help link, covers every feature from project creation to hypothesis generation.
Drag and drop in hypothesis trees — rearrange people by dragging them.
No-shared-DNA detection — automatically recognized when two tested people don't share any DNA.
Improved
Cleaner trees — unnecessary placeholder individuals are removed automatically.
Fixed
More reliable scoring — probability scores stay accurate when you regenerate.
February 10 Update
New
POI Ancestry View — a new read-only tree view showing a Person of Interest's known ancestors from the main family tree. Use it as a reference while evaluating hypotheses — no overlays, just the documented lineage.
February 8 Update
New
Cross-platform segment support — upload segment files from both MyHeritage and FTDNA for the same people. Kinprove now handles chromosome-level deduplication automatically — first platform wins per chromosome, so upload MyHeritage first for denser coverage, then add FTDNA for X chromosome data.
February 5 Update
Improved
DNA Evidence Review improvements — the review now supports multi-path relationships and shows contextual notes explaining unusual patterns like endogamy adjustments, X-DNA path issues, and consanguinity. Filter by individual or status to focus on pairs that need attention.
January 28 Update
New
Hypothesis tree filters — toggle individual hypotheses on and off in the hypothesis tree view, and show or hide non-ancestral spouses for a cleaner layout.
Rescore or recalculate — rescore existing hypotheses after uploading new DNA data, or recalculate from scratch after changing the family tree.