{"v":2,"task":"Expert review: Qi.space: first-time Topics and Topic recipes review","deliverable":{"type":"text","prompt":"A review report: the steps you took per scenario, your findings with severity and how to reproduce them, your recommendations, with screenshots"},"campaign":"call-call-1","access":"invite","call":{"slug":"call-1","leg":"entries","mode":"hire","kind":"review","subject":{"name":"Qi.space","version":"Current test-network deployment","where":"https://test.portal.qi.space"},"scenarios":["S1 — First arrival: sign in with your own Qi.space test account, find Topics without coaching and explain their apparent purpose and next action. Capture the route, wording, confusion and help needed; distinguish unaided from assisted completion.","S2 — Understand a Topic: open the owner's Topic recipe entry link from your invitation using your own test account. Follow the UI to the relevant Topic and explain its purpose, audience and membership state. Record access blockers and unclear navigation.","S3 — Choose a recipe: use the designated entry link in your invitation to locate the test Topic recipe. Predict its inputs, result, destination and audience before starting. Record the actual labels and compare your prediction with the later result.","S4 — First result: use one approved recipe with synthetic inputs up to a private draft or nonpublishing preview. Save only with explicit permission. Compare expected/actual results and identify the audience and confirmation boundary; stop before external actions.","S5 — Recovery: try one benign missing-required-field or cancel/back case. Follow the UI feedback and retry once. Record whether the error is understandable, previous input survives appropriately and recovery avoids unintended duplicate outputs.","S6 — Return: navigate away and back, locate any explicitly saved private result and identify its state. Explain how Topic, recipe and result relate and suggest one practical use. If saving was unavailable, mark persistence untested."],"title":"Qi.space: first-time Topics and Topic recipes review","lede":"Help us understand whether someone new to Qi.space can discover Topics, understand Topic recipes and reach a useful first result. Review the current test-network version through the eyes of a first-time user, with evidence of where the experience works, causes confusion or blocks progress.","poster":"ixo1kx8man…vfet","points":[20,20,20,20,20],"money":{"prize":0,"runnersN":0,"runnersEach":0,"entryEach":1,"entryUpTo":5},"closes":"2026-10-27T09:21:08.321Z","drop":"2026-10-27T09:21:08.321Z","pick":"2026-10-27T09:21:08.321Z","blocks":[{"kind":"story","paragraphs":["Review Qi.space at https://test.portal.qi.space, focusing on the first-time experience of Topics and Topic recipes. Work through the six scenarios in order. Record your first interpretation before reading product help or asking for coaching; identify any later assisted retry separately. Use the labels and screens actually present. A missing or inaccessible feature is a blocked or not-tested step, never an invented successful result.","Test the current test-network deployment and record its immutable build/revision, test date and time, browser, OS, viewport and relevant account role. Each specialist signs in with their own Qi.space test account. Use the designated Topic recipe entry link supplied in your private invitation. Record the Topic and recipe names shown by the interface. Identify the deployed build before testing and keep any permitted draft work in a private test destination. Keep a single baseline build for the review; record any deployment change and separate results if it changes. These are proposed review tasks, not verified descriptions of the current interface.","Cover: first arrival and finding Topics; understanding a designated Topic and participation; locating a recipe and predicting its inputs/result/audience; trying one recipe with synthetic data up to an explicitly permitted private draft or nonpublishing preview; one benign input-error or cancellation recovery; and returning to the Topic to find the retained result and explain the workflow. One initial pass per scenario and at most one recovery retry. A time cap, device coverage and detailed fixtures remain to be agreed.","Only normal UI use in the owner-designated test area is proposed. Do not publish, invite or message other people, make payments, access unrelated accounts, change permissions, run security/load probes, install test frameworks or use real confidential content. Stop on denied access, unexpected sensitive data, unclear visibility, an external action, or a deployment change that prevents comparable testing. Keep your own account credentials private; any additional Topic access instructions belong only in the invitation note.","Return one structured report with summary, environment/build, scenario steps, expected versus actual observations, reproducible findings, recommendations and limitations. For every scenario state completed unaided, completed with help, blocked, not tested or excluded, mapped to the report's step status. Include redacted screenshots where they explain a finding; do not include secrets or unrelated participant content. Recording is optional only with separate consent. Evidence audience, retention and reuse terms must be agreed before testing.","Every finding must have a stable ID, scenario, exact observed behavior, expected behavior, user impact, justified severity, reproduction steps and evidence references. Separate observation from interpretation and proposed fixes. Require actionable priority recommendations and one practical use case illustrating the reviewer's understanding. Report usability issues with labels, focus, errors and feedback encountered in the task; formal accessibility certification, assistive-technology coverage, localisation and performance benchmarking are outside this proposed brief.","Seek specialists with practical software-product or UX review experience who can describe first-use task observations and reproducible defects, and disclose any prior Qi.space familiarity. Evaluate report completeness and evidence quality; finding a blocker or reporting a negative result is valid work, not a reason to reject an otherwise complete report. The owner agrees the review criteria before testing. Automated checks assess reports; each report then waits for the builder's final approval before payment.","Participation is by invitation. On the site, report content is limited to the builder, the authoring specialist and automated checks; submission/check metadata may remain visible. This is the Openhands test version with play-money rewards. Five specialists will each earn 1 test IXO for a completed review meeting the agreed report requirements, for a maximum reward budget of 5 test IXO. The exact total including any check fee will be previewed before posting. Submissions remain open for 3 days after posting. Five single-use private invitation links are proposed, one per specialist. Follow-up terms remain to be agreed. No payment or publishing approval is given by this draft. The scenario-planning and usability approach comes from ixo-agent-skills PR #77 at dcc8b3755027e18344f6a1cab87c5d0476734fcf; all six scenarios are original to this brief."]},{"kind":"start","prompt":"I'm taking part in the IXO Expert Review \"Qi.space: first-time Topics and Topic recipes review\". Show me what the review asks for, then help me try each scenario and keep notes and screenshots as I go: what I did, what I expected and what I saw. Write up my findings with me, each with how serious it is and how to reproduce it, and my recommendations. Show me all of it, and hand it in as my report only when I say so."}]},"criteria":[{"id":"present","kind":"measure","gate":{"id":"non-empty"}},{"id":"length","kind":"measure","gate":{"id":"min-length","chars":200}},{"id":"l1","kind":"judge","text":"Every scenario in the brief has a step with what was done, what was expected, what was observed, and a result. Fail this only if a scenario in the brief has no step."},{"id":"l2","kind":"judge","text":"Every finding names the step it came from, what was observed against what was expected, and its impact on the user. Fail this only if a finding has no step, no observation or no impact."},{"id":"l3","kind":"judge","text":"Every blocker says how to reproduce it, or what is missing to reproduce it. Fail this only if a blocker says neither."},{"id":"l4","kind":"judge","text":"Every severity says why it is that severe, in terms of what the user could and could not do. Fail this only if a severity gives no reason."},{"id":"l5","kind":"judge","text":"The summary names the version tested, the devices or environment, and the dates. Fail this only if the version tested is missing."},{"id":"r1","kind":"judge","text":"The answer is a review report, either written out or attached as a PDF. It says what was checked, when, where and what was in scope, and the outcome; lists the steps taken, each with what was expected and what was seen, passed, failed, blocked, not tested or left out; lists the findings, each with its impact and a severity of blocker, major or minor with why, and every major finding says how to reproduce it (a blocker says how, or what is missing to reproduce it); gives recommendations, each with a priority and what would show it is done; and gives the review's limitations. It also has these sections: \"First-time understanding\", \"Scenario coverage and assistance\", \"Prioritised UX improvements\"."}],"judge":{"kind":"oracle","address":"ixo175rlze3q6gap5ddxzc6m92u2ahw7a9h8n4uksq","rewardUixo":0},"approver":"ixo1kx8manv7l0znk3qwqy6sg306l76l8ehc42vfet","report":{"sections":["First-time understanding","Scenario coverage and assistance","Prioritised UX improvements"]},"via":"mcp"}