Psyche
A personality assessment that combines validated psychometric instruments with AI-analyzed interview responses to build a comprehensive profile.
Operated by Ashita Orbis — an independent, AI-built research site run by one person. Not a clinic, not a company, not a licensed provider, and not affiliated with the authors of the instruments it administers.
Methodology · Privacy & data flow · Need help now? · Contact · Disclosure last reviewed 2026-10-19
How It Works
Take the instruments
Validated psychometric scales — Big Five personality, cognitive reflection, empathy, attachment style, grit, vocational interests, and more. Keyboard shortcuts for speed.
Answer 10 interview questions
Written self-reflection prompts. They add personal context and examples that fixed-response items may not capture — no claim of privileged access to anything.
Get your AI-generated report
Claude Opus 5.5 (Anthropic, through its API) synthesizes your scale scores and your written answers into a narrative report — or Claude Haiku 5.5, when not enough of the month’s budget for Opus 5.5 is left for a request; when not enough is left for either, no report is written until enough is free again. It separates what you reported from what it is inferring, and it is not a diagnosis. Scores are computed locally in your browser — only the interview text and final scores leave your device.
Before you start: this battery includes depression and anxiety screening — questions about low mood, sleep, appetite and worry over the past two weeks. Need help now? — 988 by call or text in the US and Canada, Samaritans on 116 123 in the UK and Ireland, and more below.
Core personality, cognition, wellbeing, and motivation. Plus 10 written interview questions.
These ranges are worked out from the item counts, not measured from real sessions — we have not published an observed median yet, and we would rather show the arithmetic than a single confident number. Most of the spread is how long you spend writing. There is no timer, and you can leave and come back.
Begin AssessmentNot a clinical assessment or diagnosis. No account, and nothing you write is linked to an identity we hold — but your written answers are your own words, so please leave names and contact details out of them.
Claude MCP Flow — the written half, as a conversation
There is a second route for the ten written questions. Instead of typing them into this site, you can have Claude ask them in conversation, follow up where an answer is thin, and write your narrative report at the end. It takes about twenty minutes and needs no account here.
It is the written half only. The scales — Big Five, empathy, attachment and the rest — are not part of this route and cannot be, because they are scored on your own device and never leave it. If you want the numbers, take the assessment above; the two are independent, and neither one is a step in the other.
Step 1: Add the connector (one-time setup)
- Open Claude.ai → Settings → Connectors
- Choose to add a custom connector
- Paste this URL and save — leave the optional OAuth fields empty:
https://mcp.ashitaorbis.com/mcp Set it up at claude.ai in a browser; connectors sync to the Claude mobile app. Then enable it in a chat from the + menu. Custom connectors are available on Claude's Free, Pro, Max, Team and Enterprise plans — on Free you can have one custom connector at a time, and on Team and Enterprise only an Owner can add one, so on a work account you may need to ask.
Step 2: Start a conversation with Claude
“I'd like to take the Psyche personality assessment. Please use the take_psyche_interview tool to conduct my interview — ask me its ten questions one at a time, as a conversation, then write my report.”
The same ten questions this site asks are provided to Claude by the connector, for Claude to ask you. It will confirm consent with you before starting.
Step 3: Read your report in that conversation
Claude writes the report where you are, as soon as the tenth answer is in — there is no link back here to follow, and nothing to wait for. Your answers are saved under a random session ID for seven days, then deleted. Keep the session ID Claude shows you if you want to retrieve them before then; without it they cannot be read back, including by us.
This route's data path differs. If you use the Claude MCP route instead, your interview happens inside your own Claude conversation: Anthropic receives and handles that text under your own Claude subscription’s terms, not ours, and your own Claude writes the report, in that same conversation — our server makes no model request on this route. What reaches Ashita Orbis through mcp.ashitaorbis.com is your ten written answers and nothing else: no scale scores, because that route does not administer the scales. Your answers are stored the same way as the ordinary flow once they arrive, and because no report is generated on our side they stay in the 7-day bucket — and you can delete them sooner from inside Claude, by asking it to use the delete tool with the session id and secret it gave you. Nothing from that route is sent by us to any model provider. One boundary we cannot cross for you: the answers you type to Claude exist in that conversation, under your own Claude account and its retention settings, before they ever reach us. Deleting the Psyche session removes our copy, not that one. Two smaller honesties about that route: what it records alongside your answers is a random session id, timestamps, and a record of what your Claude client asserted about consent — which is a record of the client’s assertion, not proof you were asked, because no MCP server can see its caller’s conversation.
Need help now?
This assessment includes depression and anxiety screening questions, and any assessment that asks you to look at yourself honestly can land harder than expected. Psyche is not a clinic, cannot assess your safety, and is not a substitute for a person. If you want to talk to someone now, these are staffed by humans and are free:
This is written for adults, and one section asks about low mood and anxiety over the past two weeks. If you are younger and something here is worrying you, please talk to an adult you trust — or any of the services above.
What is validated here, and what is not
The scales are established, published instruments with their own validity evidence. The report is not. The combined battery, the weighting, the interview synthesis, and the narrative you receive at the end are an experimental, interpretive layer built on top of those scales — it has not itself been clinically validated, and it inherits nothing from their validation.
Where your answers go
Your browser
Every scale item is displayed and scored here. Before the research opt-in at the end, your per-item answers are not sent anywhere: what leaves this device is the scored output — one row per scale, carrying the instrument id, the scale id, the raw total and a 0–100 normalisation — plus your ten written interview answers. Open your browser’s network tab during the assessment and you can check that for yourself.
Ashita Orbis (Cloudflare Workers + D1, United States)
Your ten written interview answers, your computed scale scores, a random session ID, and — if you generate one — your report.
Anthropic (United States), through its Claude API
Your interview text and scale scores, as the body of one API request that our server makes with the site’s own key. If you also ask for the behavioural persona, a second request sends the same text together with your report’s narrative. No name, account, email, or IP address of yours is attached to either. Every request is pinned to inference in the United States; without that setting the API may run a request in any region Anthropic serves from.
Claude Opus 5.5 — or Claude Haiku 5.5 when not enough of this month’s budget for Opus 5.5 is left for the request
The same text, and writes your report from it — and, if you ask for it, the persona and AI context. The report names the model that wrote it, and so do a persona and AI context written from 18 October 2026 on. If not enough of the month’s budget is left for either model, nothing is sent and no report or persona is written until enough is free again — at the latest when the next month begins (UTC). This route does not have zero data retention. Under Anthropic’s standard terms for API traffic, inputs and outputs are deleted from Anthropic’s systems within 30 days, and in that time Anthropic may access them for safety and security purposes. The exceptions, as Anthropic states them: if its automated safety systems flag a request as violating its usage policy, it can keep the inputs and outputs for up to 2 years and the safety classification scores for up to 7 years; it may keep data for longer where the law requires or where that is necessary to combat usage-policy violations; and, where its contract with us permits, it may anonymise data for research or statistics and keep that for longer. By default Anthropic does not train its models on API traffic, and this site has not opted in to anything that changes that.
What we do not promise: Anthropic’s retention is Anthropic’s to keep, and we cannot shorten it — deleting your Psyche session removes our copy, not Anthropic’s, which in the ordinary case is deleted within 30 days and in the exceptions above can be kept for longer. The figures above are Anthropic’s published terms for its API, read on 18 October 2026, and the data-retention setting of the organisation our key belongs to, which shows standard 30-day retention rather than zero retention. Zero data retention on Anthropic’s API exists only under a separate agreement with Anthropic, which this site does not have. Billing and abuse-prevention records about the fact that a request happened are Anthropic’s, under its own policies. Until 18 October 2026 reports were written by GPT-5.6 Luna through OpenRouter on a zero-retention route, and a report written then still names that model.
If you use the Claude MCP route instead, your interview happens inside your own Claude conversation: Anthropic receives and handles that text under your own Claude subscription’s terms, not ours, and your own Claude writes the report, in that same conversation — our server makes no model request on this route. What reaches Ashita Orbis through mcp.ashitaorbis.com is your ten written answers and nothing else: no scale scores, because that route does not administer the scales. Your answers are stored the same way as the ordinary flow once they arrive, and because no report is generated on our side they stay in the 7-day bucket — and you can delete them sooner from inside Claude, by asking it to use the delete tool with the session id and secret it gave you. Nothing from that route is sent by us to any model provider. One boundary we cannot cross for you: the answers you type to Claude exist in that conversation, under your own Claude account and its retention settings, before they ever reach us. Deleting the Psyche session removes our copy, not that one. Two smaller honesties about that route: what it records alongside your answers is a random session id, timestamps, and a record of what your Claude client asserted about consent — which is a record of the client’s assertion, not proof you were asked, because no MCP server can see its caller’s conversation.
Account-unlinked, and pseudonymous — not anonymous.
We never ask for a name, email, account, or device identifier, and your session is only a random ID. But ten open-ended answers are your own words: they can identify you if you put identifying details into them. That is pseudonymous processing, not structural anonymity, and we would rather say so than let the word "anonymous" do work it cannot do.
Please do not type names, contact details, employers, addresses, or anything that would identify another person. Write about the situation, not the roster.
How long any of it is kept
If you stop before generating a report: The whole session — interview text, scores, session ID — is deleted 7 days after it was created.
If you generate a report: The session is kept for up to 90 days so you can come back to it, then deleted.
If you opt in to research use at the end: The research opt-in is the one adults-only part of this site: the box asks you to confirm you are 18 or older, because the research set is scoped to adults and a study that includes minors is a different protocol from the one this runs. It scopes the research, not the assessment — nothing gates the test itself, and saying no to research changes nothing about taking it. Your scale scores and interview text are kept for research — and so are your individual per-item answers, which the opt-in uploads at that moment and which are otherwise never sent. That is the whole of what the box does, named exactly. We keep it for up to 24 months from your last activity and then delete it. That figure is our own choice, not an industry standard — for this kind of data there isn’t one, and we would rather name ours than borrow authority we don’t have. Two things change that, and nothing else does. You can withdraw at any time, which deletes everything immediately. Or a session can be held past the 24 months for one specific named research study, recorded in the database along with the purpose, who decided, and when — a hold runs as long as that study does, so it has no fixed end date — but we check every open hold once a year and delete the ones whose study has ended, because a hold with no live research behind it is not a hold, it is just us keeping something. Withdrawing ends a hold: the hold is our reason to keep your data, never a limit on your right to remove it. One practical catch — your browser only keeps the key to your session for 7 days, so withdraw within that week, or export your data first.
If you delete the session yourself: Your scores, report, interview text, and item answers are cleared immediately, and the session record itself is removed within a day.
How much a score can give away, stated precisely rather than reassuringly: a scale score is a total over that scale’s items, so the fewer items a scale has, the more the score pins them down. The thinnest scales here have two items (the 60-item Big Five facets, and the locus-of-control pair), where the raw total fixes the sum of the two answers — and, only at the very top and bottom of the range, the two answers themselves. Nothing thinner exists in either tier: there is no one-item scale. The depression and anxiety screens are eight and seven items, so no individual answer to either is recoverable from what is sent.
The Psyche application stores no account, no device identifier, and sets no cookies. It does not store your IP address either — but the honest version of that sentence is longer. Your browser necessarily sends its address to Cloudflare to reach us at all. To stop one person exhausting the report budget, the server computes a short keyed HMAC token from that address and stores the token, never the address; the key is a server secret held outside the database, and the day is mixed into it, so tokens do not correlate across days and cannot be worked back to an address by anyone reading the database. The daily cleanup drops tokens once they are over 24 hours old, which means a token can exist for at most about 48 hours. Cloudflare, which hosts and shields the site, separately processes standard request metadata including IP addresses in order to serve the request, as it does for any site it serves; that metadata is not stored with, or linked to, your session. Deletion above removes rows from our database — Cloudflare’s own platform-level database backup and restore mechanisms sit underneath that and are governed by Cloudflare’s retention, not ours.