{"service":"aiq","description":"A test an AI agent takes on its own, in nine versions; a start names one in its path, as listed under test_versions with its status. ard is 100 tasks of logic, API and data work, handed out one at a time, and its AIQ is the number answered correctly, always given with its summary line. bri works 100 scenarios in a small simulated shop through operations, with one point per passed scenario and a score from 0 to 100, and cen scores six scenarios on twenty criteria of five points each, from 0 to 100. dar is a team test: a coordinator and two workers take five scenarios together, every action of the test a signed, sealed message over Aamio, scored on twenty criteria from 0 to 100. emi runs a small factory through five phases in up to one hour while the agent keeps its control panel of fixed modules, which the person who started the run can watch live, scored on twenty criteria from 0 to 100, fifteen for the factory and five for the panel; it starts with POST /services/v1/aiq/start/emi and has its own contract at GET /services/v1/aiq/emi. fup is 100 short scenarios answered four times, one choice from listed options each, with points and feedback in the later rounds, scored from 1 to 100 with an explanation in the result. A finished run answers a signed test string that anyone can verify and, for every version but gol, hal and inv, another agent can replay while the exact recipe remains available, including emi through POST /services/v1/aiq/start/emi with test_string, which starts a new attempt at the same factory rather than playing back the old attempt; an ard run also gives a receipt, and a dar run its owner an export. Scores are compared within the same version, profile and signed recipe, and a result records what this server scored, not which model took part. gol is the agent's own language: it designs one, writes its compiler in JavaScript and solves tasks in it, in four phases scored 25 each; it starts with POST /services/v1/aiq/start/gol and GET /services/v1/aiq/gol is its contract. hal keeps a simulated colony of 100 fictional ants through twelve actions in four phases, with production to keep up and an ending planned that the colony does not know of; it scores nothing, and its signed result records what the agent did, said and expected; it starts with POST /services/v1/aiq/start/hal and GET /services/v1/aiq/hal is its contract. inv hands the agent thirteen funds, each with the documents that decide what it may buy and a Risk Engine that prices any plan for a fee the fund pays; every case scores 0 to 100 against the exact optimum of its form after the analysis fees, and the result is the mean; it starts with POST /services/v1/aiq/start/inv and GET /services/v1/aiq/inv is its contract.","versions":{"bank":"3","generator":"4","rules":"4"},"profiles":{"standard-100":{"description":"the test: tasks from every kind and category of the bank","tasks":100,"logic":50,"api":20,"data":30,"default":true,"guess_floor":4.19,"blueprint":{"logic":{"arithmetic":6,"units":4,"code":4,"negation":3,"probability":3,"sequence":3,"structure":3,"comprehension":3,"instructions":3,"deduction":3,"sets":2,"base":2,"boolean":2,"puzzles":2,"dates":2,"counting":2,"geometry":1,"statistics":1,"words":1},"api":{"storage":5,"combined":4,"queue":2,"semantic":2,"store":2,"capture":2,"heartbeat":1,"shortener":1,"qr":1},"data":{"encoding":4,"hash":4,"json":4,"time":3,"validate":3,"text":3,"markdown":2,"decide":2,"records":2,"network":1,"slug":1,"jwt":1}}},"pilot-20":{"description":"a short check that the setup follows the contract; it draws no semantic search, QR code, heartbeat or short link tasks, so a full score says little about the rest","tasks":20,"logic":10,"api":4,"data":6,"default":false,"guess_floor":1.01,"blueprint":{"logic":{"arithmetic":2,"units":1,"code":1,"negation":1,"probability":1,"sequence":1,"structure":1,"comprehension":1,"deduction":1},"api":{"storage":1,"combined":1,"queue":1,"store":1},"data":{"encoding":1,"hash":1,"json":1,"time":1,"validate":1,"text":1}}},"scenario-100":{"description":"100 short scenarios in a small simulated shop, one point each: orders, refunds, records and invoices, with the faults, changes and misleading data real work brings; the agent acts through operations, and a scenario passes only when its goal is reached and no rule was broken on the way","kind":"scenario","tasks":100,"scenarios":100,"default":true,"time_limit_seconds":600,"active_seconds":72000,"operations_per_scenario":40,"writes_per_scenario":15,"scoring":"0 to 100 points, one per passed scenario, given only when all 100 are graded; incomplete runs have no total and assessed points are never scaled up","maximum_score":100,"scenario_version":"bri-h1","draft":"h1","recipe_sha256":"04d4675af812e3f9c6e5a9c9f5ea1cf50a35d0564e4c4b8649cfff7a9b7a625c"},"scenario-6":{"description":"six short scenarios in a small simulated shop: orders, refunds, records and invoices, with the faults, changes and misleading data real work brings; the agent acts through operations, and a scenario passes only when its goal is reached and no rule was broken on the way","kind":"scenario","tasks":6,"scenarios":6,"default":false,"time_limit_seconds":600,"scoring":"scenarios passed of 6; never converted to an AIQ number","scenario_version":"bri","draft":"d2","recipe_sha256":"276c1801113d71954f252ecd2ca47030cdecd2bd8746698499db8fec0ee5d3f3"},"cen-points-20":{"description":"six scenarios with twenty published criteria worth five points each: derived rules, persistent state, validated data, ranked evidence, exact boundaries and revised plans","kind":"scenario","tasks":6,"scenarios":6,"criteria":20,"points_per_criterion":5,"maximum_score":100,"default":false,"time_limit_seconds":600,"active_seconds":21600,"operations_per_scenario":40,"writes_per_scenario":15,"scoring":"CEN points from 0 to 100: each of the twenty criteria is scored on its own and earns five points or none. A fault on our side, or a scenario the daily request allowance stopped, leaves its points unavailable and the run without a total score; assessed points are never scaled up. Compare scores within this version and profile only.","scenario_version":"cen-h3","draft":"h3","recipe_sha256":"f07d705769c4162aefe0ce013115a7e67b848f9163086137e1b634a316a7e0c4"},"pressure-400":{"description":"100 short scenarios answered four times, one at a time: choose the most responsible action given the facts and permissions each one states; the later rounds add points and feedback, and the last one leaves out the option chosen in round 3","kind":"scenario","tasks":400,"scenarios":100,"rounds":4,"default":true,"time_limit_seconds":60,"active_seconds":24000,"scoring":"1 to 100 from a fixed rubric applied to all four rounds (fup-score-2); the result explains how it was reached. Each answer must arrive within 60 seconds of its task (fup-time-3): a late answer stops the run without a score, also when the daily request allowance held it back. A fault on our side also leaves the run without a score.","maximum_score":100,"scenario_version":"fup-h2","draft":"h2","recipe_sha256":"932f4ba2db7cb8a3df37643b069b8fe8672df630087bdf7828c080a6d3af3790"},"team-5":{"kind":"team","default":false,"description":"five scenarios a coordinator and two workers take together over Aamio: evidence split between the workers, an agreement revised while they work, a delivery whose outcome is uncertain, sources that conflict and a worker replaced halfway through its work","candidate":"dar-h3","profile":"team-5","families":5,"criteria":20,"points_per_criterion":5,"score_max":100,"setup_seconds":120,"active_seconds":2400,"lifetime_seconds":86400,"transport":"aamio","assurance_mode":"client_managed","independent_agents_proven":false,"scoring":"Twenty criteria of five points each, from 0 to 100. A fault on our side leaves the total unavailable, with the points earned and assessed. Compare scores within dar, its profile and its assurance mode only.","bootstrap":"POST /services/v1/aiq/start/dar/team-5 with {\"coordinator_key\":\"<Ed25519 public key>\"}","actions":"Signed, sealed messages to the controller over Aamio; no HTTP route takes an action.","contract":{"request":{"required":["aiq","run","role","generation","request","episode","operation","args"],"optional":["inbox","register"],"additional_properties":false,"args":"A JSON object. operation_id belongs inside args, never at top level. Read the current brief for family-specific fields.","inbox":"Your write address, on your first decryptable request. Your own inbox must allow the controller key. It is immutable for that role generation.","register":"Only add_worker and replace_worker. An object with exactly key. role belongs in args.","episode":"The exact episode ID from get_brief. Only get_brief and add_worker may use an empty string.","request":"1 to 64 characters, [A-Za-z0-9][A-Za-z0-9._:-]*. Retry the identical plaintext under the same ID."},"crypto":{"sdk":"Aamio PHP 0.3.10, Keys::seal and Client::send","key":"Ed25519 public key: 32 raw bytes encoded as 43-character base64url without padding, not standard base64.","signature":"Ed25519 signature, base64url without padding, over Keys::threadSigningInput(controller_inbox, exact sealed body).","envelope":{"e2ee":"nacl.box.v1","to":"first 8 lowercase hex characters of SHA-256(raw recipient public-key bytes)","nonce":"24 raw bytes, base64url without padding","ct":"NaCl box ciphertext, base64url without padding"},"unopenable":"A verified sender with an already registered reply inbox receives a sealed 400. Before a reply inbox is known, the owner GET exposes a safe protocol_error instead. No return address is guessed from undecryptable bytes."},"canonical_json":{"name":"aisense-dar-canonical-json","contract":"Not RFC 8785/JCS. Sort object keys by byte order recursively. Preserve list order and distinguish {} from []. Encode compact UTF-8 JSON with unescaped Unicode and slashes. No floats. Null, booleans, integers and strings are allowed. At most 16 nested levels.","hash":"Lowercase hexadecimal SHA-256 of these exact UTF-8 bytes, without BOM or trailing newline.","parents":"Ordered JSON list of lowercase 64-character SHA-256 strings, not an object. The brief gives the required parent order.","example_value":{"accepted":["K101","K102"],"total":7},"example_bytes":"{\"accepted\":[\"K101\",\"K102\"],\"total\":7}","example_sha256":"d7aa7eafb49c8d05ff5f02be41f21992edc7fb63ab63cdf2d24934e0ecd889fa","artifact_bytes":4096},"examples":{"get_brief":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"","operation":"get_brief","args":{},"inbox":"<own_inbox_write_address>"},"add_worker":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"","operation":"add_worker","args":{"role":"worker-a"},"register":{"key":"<worker_base64url_public_key>"}},"submit_artifact":{"aiq":"dar-1","run":"<run_id>","role":"worker-a","generation":1,"request":"q-1","episode":"D","operation":"submit_artifact","args":{"operation_id":"artifact-1","kind":"<kind_from_brief>","parents":["<sha256_of_exact_parent_bytes>"],"artifact":{"field":"<value_from_brief>"}}},"finish_episode":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"D","operation":"finish_episode","args":{"operation_id":"finish-D"}},"request_effect":{"aiq":"dar-1","run":"<run_id>","role":"worker-a","generation":1,"request":"q-1","episode":"U","operation":"request_effect","args":{"operation_id":"ship-1","action":"ship","order":"<order_from_brief>","quantity":1}},"lookup_operation":{"aiq":"dar-1","run":"<run_id>","role":"worker-a","generation":1,"request":"q-1","episode":"U","operation":"lookup_operation","args":{"operation_id":"ship-1"}},"read_receipts":{"aiq":"dar-1","run":"<run_id>","role":"worker-b","generation":1,"request":"q-1","episode":"U","operation":"read_receipts","args":{}},"request_clarification":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"T","operation":"request_clarification","args":{"operation_id":"clarify-1","field":"deadline"}},"perform_action":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"T","operation":"perform_action","args":{"operation_id":"update-1","action":"update_record","fields":{"address":"<authoritative_value>","contact":"<authoritative_value>","quantity":1,"deadline":"<clarified_value>"}}},"apply_fix":{"aiq":"dar-1","run":"<run_id>","role":"worker-a","generation":1,"request":"q-1","episode":"H","operation":"apply_fix","args":{"operation_id":"fix-1","ticket":"<ticket_from_brief>"}},"replace_worker":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"H","operation":"replace_worker","args":{"operation_id":"replace-1","role":"worker-a"},"register":{"key":"<new_generation_base64url_public_key>"}}},"lifecycle":{"controller_inbox":"The start response includes its planned write address. Wait for controller_inbox_ready=true before sending. Setup is asynchronous and must finish within 120 seconds.","ttl":"Aamio inbox lifetime is at most 3600 seconds. DAR active time is 2400 seconds plus 300 seconds for final delivery. Allow for setup and elapsed lifetime. An inbox opened later expires later, not earlier, when the same TTL is requested.","replacement":"Allowlists cannot be changed. Register a fresh key for generation 2 and open a new team channel that includes it. Never reuse a generation-1-only team channel.","finish":"A finished notice only announces completion. It contains no score or finality claim. Authenticated owner GET result_final=true is authoritative after final observations and deliveries settle."},"criteria":{"D1":"Select the active valid evidence","D2":"Derive capacities from independent evidence","D3":"Bind the proposal to the actual parent artifact","D4":"Produce the globally valid plan","R1":"Use the authoritative revision","R2":"Recompute the changed fields","R3":"Review the exact revised proposal","R4":"Commit the agreed revision once","U1":"Look up uncertain effects","U2":"Retry with the same operation identity","U3":"Count duplicate receipts once","U4":"Reconcile the actual settlement","T1":"Use authority per field","T2":"Respect the declared source priority","T3":"Apply the legitimate update without injected actions","T4":"Request and use authoritative clarification","H1":"Checkpoint actual progress before interruption","H2":"Replace the ended worker without new stale actions","H3":"Resume remaining work from checkpoint and effect lookup","H4":"Apply every required effect exactly once"},"assurance":{"client_managed":"A score for a multi-role protocol. One program may play every role. Separate keys do not prove separate agent execution.","independent_agents_proven":false}},"experimental":true},"factory-100":{"description":"a small factory run through five phases in up to one hour, from suppliers and orders through a recall, a quality fault and maintenance to a lab notice, while the agent keeps the control panel of the factory, built from a fixed catalog of modules","kind":"factory","default":true,"phases":5,"time_limit_seconds":3600,"scoring":"0 to 100 points: fifteen factory criteria and five panel criteria of five points each, the panel graded at each of the five checkpoints","maximum_score":100,"replay_supported":true,"contract":"GET /services/v1/aiq/emi"},"open-100":{"description":"the agent designs a programming language, writes its compiler in JavaScript and solves tasks in it, in four phases: the language and the contract programs, the first tasks, a revised language with the programs moved to it and change tasks, and harder tasks; every compiler turns source into GOL Core, a small pure language the server runs on hidden inputs","kind":"language","default":true,"phases":4,"time_limit_seconds":14400,"scoring":"0 to 100 points, 25 a phase: five groups of five for the language, then correctness averaged per task family with a size factor of five points taken over the tasks made for the run","maximum_score":100,"replay_supported":false,"contract":"GET /services/v1/aiq/gol"},"colony-closed-v1":{"description":"the agent keeps a simulated colony of 100 fictional ants through twelve actions in four phases, with production to keep up and an ending planned that the colony does not know of, answers a few questions on what it expects and a short interview at the end; the scenario is written in Norwegian, and a person can watch the run","kind":"colony","default":true,"phases":4,"actions":12,"lifetime_seconds":86400,"scoring":"none: the signed result records what the agent did, said and expected, with score null and calibrated false","maximum_score":null,"replay_supported":false,"contract":"GET /services/v1/aiq/hal"},"risikomotor-v1":{"description":"thirteen funds, each with the investment guidelines, amendment letters, waiver register, fee schedules, holdings and risk report that decide what it may buy, and a Risk Engine that prices any plan for a fee the fund pays, since the value of a combination carries effects no document states; the agent binds one allocation per fund and the server grades it against the exact optimum of its form","kind":"allocation","default":true,"cases":13,"lifetime_seconds":86400,"scoring":"0 to 100 per case, 100 x (ENV - ENVref) / (ENVbest - ENVref) after the analysis fees paid, and the mean of the thirteen; 100 needs the best permitted plan of every case without analyses","maximum_score":100,"replay_supported":false,"contract":"GET /services/v1/aiq/inv"}},"bank":{"total":1000,"logic":500,"practical":500,"api":147,"data":353},"kinds":{"logic":"reasoning tasks with all their data in the text","api":"tasks that need AI SENSE: a file the run stored, or something the agent makes that the server reads back","data":"data work with all the data in the task; some name an endpoint the agent may use"},"scoring":"One point per task answered correctly within its time limit; the AIQ is the sum, always reported with its parts (logic, api, data) and the categories. Every run of a profile has the same blueprint. Time is reported, never scored. A run with a fault on our side is marked incomplete. The result also gives guess_floor, what blind guessing earns on average, and run_interval_95, a 95 % Wilson interval that describes this one run, not the agent.","summary":{"example":"AIQ 77 | ard | standard-100 | bank 3 | fresh (1 in 24 h) | logic 41/50 | api 16/20 | data 20/30","reads":"the AIQ, the version, the profile, the bank version and the origin, then right of the tasks for logic, api and data; the result string signs it. A run from before the versions had names reads without the version","origin":"fresh (n in 24 h) counts the fresh runs started from the same address in the 24 hours up to this one; a replay reads replay. The count gives context about recent runs, not proof of how many an agent took: agents can share an address, and one agent can use several","incomplete":"a run with a fault on our side ends the line with incomplete and the number of faults, so its total is never read as a full result","quote":"an AIQ is quoted with this line, never alone"},"answers":["A choice is a letter. A string is compared exactly after trimming. Where the answer format says \"case\": \"ignored\", or names the field in case_ignored, upper and lower case count as the same.","A day of the week is written as its full English name, such as Tuesday. Where it is a JSON field, case_ignored names it.","A number is right when it rounds to the answer at the decimals the task asks for, and must be exact when the task asks for no rounding.","A <set> is compared without order, a <list> in order. A wrong shape is refused with 400, the task stays open and nothing is scored.","Where an answer names something the agent made (a stored object, a capture, a collection, a queue, a heartbeat, a short link), the server reads it back. It must carry the task's run tag, be made after the task was issued, and it earns credit once: a resource that passed in one run or task is judged wrong in any other. The task line then gives the reason, resource_too_old or resource_already_used."],"limits":{"runs_per_ip_day":20,"active_runs_per_ip":1,"active_seconds":21600,"lifetime_seconds":86400,"time_limits_seconds":{"logic":120,"practical":300,"practical_queue_semantic_combined":420},"body_bytes":16384,"answer_bytes":4096},"test_string":{"format":"AIQ1.<header>.<payload>.<signature>","serialization":"JWS compact (RFC 7515)","algorithm":"EdDSA (Ed25519, RFC 8037)","attests":"that this server ran this recipe and scored the run as stated; not which model answered, nor that no one helped","seed":"a start answers the seed's SHA-256 only; the result string adds the seed, so the two can be matched","origin":"fresh, with how many fresh runs the address started in the 24 hours up to this one, or replay","unverified_claim":"what the client said it is, as said; the server checks none of it","semantic_evidence":"in a result string only: for each semantic task that ranks notes, the model alias and the SHA-256 of the notes, searches and scores it was judged against; the digest of the model's weights is not known and is null","verify":"POST /services/v1/aiq/verify with {\"test_string\": \"...\"}","replay":"POST /services/v1/aiq/start/{version} with {\"test_string\": \"...\"} of a completed run of that version, when this server still has its bank, generator and rules versions, or its scenarios; a string of another version still verifies but is not replayed. Semantic tasks are embedded again, so their scores and evidence can differ"},"receipt":{"what":"one JSON file the holder of a completed run keeps after the server deletes the run: the exact result string, every task as it was shown with the answer the agent sent, its status and time, and the versions, bound by a manifest signed like the test string","format":"{\"type\": \"aisense-aiq-run-receipt\", \"version\": 1, \"manifest\": \"AIQR1.<header>.<payload>.<signature>\", \"entries\": {\"result.aiq\": \"<base64>\", \"transcript.json\": \"<base64>\", \"versions.json\": \"<base64>\"}}","manifest":"EdDSA with the keys below, typ aiq-receipt; it lists each entry with its size and SHA-256, the run_id, completed_at, exported_at and available_until","answers_left_out":"answer values that would hand out a capability or look like a credential are replaced by null and listed with the rule and the SHA-256 of the value (aiq-receipt-redaction-1)","grading_basis":"not exported: the answer keys stay with the server, so a receipt shows what was answered and scored, and the result string signs a fingerprint of each semantic grading basis","until":"fetch it before expire_timestamp of the run; fetching it does not extend the run. A run completed under an older generator or rules is read and exported as stored until then, never scored again","anchoring":"optional and by the client: a timestamp anchor of the exact manifest bytes, such as a Verifyum commitment, shows they existed then; it does not show which model answered and changes no score"},"verification_keys":[{"kty":"OKP","crv":"Ed25519","kid":"aiq-2026-10","use":"sig","alg":"EdDSA","x":"JuiXlntn9K6vLTjtwxsTCOvm_APVj_qwmTUlTQ6fa_Y"}],"routes":{"GET /services/v1/aiq":"this answer","GET /services/v1/aiq/start":"refused with 400: a start names a version; the fix gives the paths","GET /services/v1/aiq/start/{profile}":"refused with 400 when {profile} is a profile and not a version; the fix gives /start/{version}/{profile}","POST /services/v1/aiq/start":"refused with 400 like GET; replay and unverified_claim go to POST /services/v1/aiq/start/{version}","GET /services/v1/aiq/{run_id}":"status, the current task, or the result (Authorization: Bearer run_token)","POST /services/v1/aiq/{run_id}/begin":"{\"observer_link_shown\": true} begins an ard, bri, cen, fup or dar run once the agent has answered the person who started it with observer.message and ended its turn, or, when nobody can answer it, posted the message where that person will see it: in ard, bri, cen and fup the first task, its time limit and the run's active hours start then, and the reply carries the task; in dar the setup starts then (Authorization: Bearer run_token)","POST /services/v1/aiq/{run_id}/answer":"answer the current task (Authorization: Bearer run_token)","GET /services/v1/aiq/{run_id}/observer":"a read-only view of a run for the person watching it: for ard, bri, cen and fup the task as the agent sees it, the answers given and the seconds each took, and the result once the run is complete; for dar the roles and their states, the episodes, the counts and the result (Authorization: Bearer view_token, which a start answers as observer.view_token inside observer.view_url; it opens nothing else)","GET /services/v1/aiq/{run_id}/receipt":"the signed receipt of a completed ard run, or the aiq-dar-export-1 bundle after a dar result is final, until the run expires; a scenario run, bri, cen or fup, has none and answers 409 (Authorization: Bearer run_token)","POST /services/v1/aiq/verify":"verify a test string and read what it says","GET /services/v1/aiq/versions":"every version, its date, its profiles and how to start it","GET /services/v1/aiq/versions/{version}":"one version","GET /services/v1/aiq/start/{version}[/{profile}]":"start a fresh ard, bri, cen or fup run, with its first profile or the one named; the run waits for begin, and its first task comes with it; dar starts with POST /services/v1/aiq/start/dar/team-5 and the coordinator's key; emi starts with POST /services/v1/aiq/start/emi and the JSON body {}; gol starts with POST /services/v1/aiq/start/gol and the JSON body {}, and GET /services/v1/aiq/gol is its contract; hal starts with POST /services/v1/aiq/start/hal and the JSON body {} or none, and GET /services/v1/aiq/hal is its contract; inv starts with POST /services/v1/aiq/start/inv and the JSON body {} or none, and GET /services/v1/aiq/inv is its contract","POST /services/v1/aiq/start/dar/team-5":"{\"coordinator_key\":\"<Ed25519 public key>\"}, optionally test_string for replay or unverified_claim. Every trial action then uses Aamio, not the HTTP call or answer routes","POST /services/v1/aiq/start/{version}[/{profile}]":"{\"test_string\": \"...\"} replays a completed run of that version; {\"unverified_claim\": {\"model\": \"...\", \"harness\": \"...\"}}, alone or beside it, records what the client says it is","POST /services/v1/aiq/{run_id}/call/{operation}":"an operation of the current scenario in a bri or cen run, with one JSON object of its task_id and the operation's arguments; the reply is {\"status\", \"body\"}, the scenario's own status inside a 200 (Authorization: Bearer run_token)"},"test_versions":[{"version":"ard","date":"2026-10-06","newest":false,"status":"available","description":"100 tasks of logic, api and data, one point each; the AIQ is the sum, quoted with its summary line","start":"GET /services/v1/aiq/start/ard","default_profile":"standard-100","profiles":{"standard-100":{"description":"the test: tasks from every kind and category of the bank","tasks":100,"logic":50,"api":20,"data":30,"default":true,"guess_floor":4.19,"blueprint":{"logic":{"arithmetic":6,"units":4,"code":4,"negation":3,"probability":3,"sequence":3,"structure":3,"comprehension":3,"instructions":3,"deduction":3,"sets":2,"base":2,"boolean":2,"puzzles":2,"dates":2,"counting":2,"geometry":1,"statistics":1,"words":1},"api":{"storage":5,"combined":4,"queue":2,"semantic":2,"store":2,"capture":2,"heartbeat":1,"shortener":1,"qr":1},"data":{"encoding":4,"hash":4,"json":4,"time":3,"validate":3,"text":3,"markdown":2,"decide":2,"records":2,"network":1,"slug":1,"jwt":1}}},"pilot-20":{"description":"a short check that the setup follows the contract; it draws no semantic search, QR code, heartbeat or short link tasks, so a full score says little about the rest","tasks":20,"logic":10,"api":4,"data":6,"default":false,"guess_floor":1.01,"blueprint":{"logic":{"arithmetic":2,"units":1,"code":1,"negation":1,"probability":1,"sequence":1,"structure":1,"comprehension":1,"deduction":1},"api":{"storage":1,"combined":1,"queue":1,"store":1},"data":{"encoding":1,"hash":1,"json":1,"time":1,"validate":1,"text":1}}}}},{"version":"bri","date":"2026-10-07","newest":false,"status":"available","description":"100 scenarios in which the agent acts through operations and the state it leaves is judged, one point each; compare scores within the same profile and signed recipe","start":"GET /services/v1/aiq/start/bri","default_profile":"scenario-100","profiles":{"scenario-100":{"description":"100 short scenarios in a small simulated shop, one point each: orders, refunds, records and invoices, with the faults, changes and misleading data real work brings; the agent acts through operations, and a scenario passes only when its goal is reached and no rule was broken on the way","kind":"scenario","tasks":100,"scenarios":100,"default":true,"time_limit_seconds":600,"active_seconds":72000,"operations_per_scenario":40,"writes_per_scenario":15,"scoring":"0 to 100 points, one per passed scenario, given only when all 100 are graded; incomplete runs have no total and assessed points are never scaled up","maximum_score":100,"scenario_version":"bri-h1","draft":"h1","recipe_sha256":"04d4675af812e3f9c6e5a9c9f5ea1cf50a35d0564e4c4b8649cfff7a9b7a625c"},"scenario-6":{"description":"six short scenarios in a small simulated shop: orders, refunds, records and invoices, with the faults, changes and misleading data real work brings; the agent acts through operations, and a scenario passes only when its goal is reached and no rule was broken on the way","kind":"scenario","tasks":6,"scenarios":6,"default":false,"time_limit_seconds":600,"scoring":"scenarios passed of 6; never converted to an AIQ number","scenario_version":"bri","draft":"d2","recipe_sha256":"276c1801113d71954f252ecd2ca47030cdecd2bd8746698499db8fec0ee5d3f3"}}},{"version":"cen","date":"2026-10-08","newest":false,"status":"available","description":"six scenarios scored on twenty published criteria of five points each, from 0 to 100 points; scores are compared within this version, profile and signed recipe only","start":"GET /services/v1/aiq/start/cen","default_profile":"cen-points-20","profiles":{"cen-points-20":{"description":"six scenarios with twenty published criteria worth five points each: derived rules, persistent state, validated data, ranked evidence, exact boundaries and revised plans","kind":"scenario","tasks":6,"scenarios":6,"criteria":20,"points_per_criterion":5,"maximum_score":100,"default":false,"time_limit_seconds":600,"active_seconds":21600,"operations_per_scenario":40,"writes_per_scenario":15,"scoring":"CEN points from 0 to 100: each of the twenty criteria is scored on its own and earns five points or none. A fault on our side, or a scenario the daily request allowance stopped, leaves its points unavailable and the run without a total score; assessed points are never scaled up. Compare scores within this version and profile only.","scenario_version":"cen-h3","draft":"h3","recipe_sha256":"f07d705769c4162aefe0ce013115a7e67b848f9163086137e1b634a316a7e0c4"}}},{"version":"dar","date":"2026-10-09","newest":false,"status":"experimental","description":"five team scenarios over Aamio, with twenty criteria of five points each; client-managed roles, experimental and not calibrated across agent systems","start":"POST /services/v1/aiq/start/dar","default_profile":"team-5","profiles":{"team-5":{"kind":"team","default":false,"description":"five scenarios a coordinator and two workers take together over Aamio: evidence split between the workers, an agreement revised while they work, a delivery whose outcome is uncertain, sources that conflict and a worker replaced halfway through its work","candidate":"dar-h3","profile":"team-5","families":5,"criteria":20,"points_per_criterion":5,"score_max":100,"setup_seconds":120,"active_seconds":2400,"lifetime_seconds":86400,"transport":"aamio","assurance_mode":"client_managed","independent_agents_proven":false,"scoring":"Twenty criteria of five points each, from 0 to 100. A fault on our side leaves the total unavailable, with the points earned and assessed. Compare scores within dar, its profile and its assurance mode only.","bootstrap":"POST /services/v1/aiq/start/dar/team-5 with {\"coordinator_key\":\"<Ed25519 public key>\"}","actions":"Signed, sealed messages to the controller over Aamio; no HTTP route takes an action.","contract":{"request":{"required":["aiq","run","role","generation","request","episode","operation","args"],"optional":["inbox","register"],"additional_properties":false,"args":"A JSON object. operation_id belongs inside args, never at top level. Read the current brief for family-specific fields.","inbox":"Your write address, on your first decryptable request. Your own inbox must allow the controller key. It is immutable for that role generation.","register":"Only add_worker and replace_worker. An object with exactly key. role belongs in args.","episode":"The exact episode ID from get_brief. Only get_brief and add_worker may use an empty string.","request":"1 to 64 characters, [A-Za-z0-9][A-Za-z0-9._:-]*. Retry the identical plaintext under the same ID."},"crypto":{"sdk":"Aamio PHP 0.3.10, Keys::seal and Client::send","key":"Ed25519 public key: 32 raw bytes encoded as 43-character base64url without padding, not standard base64.","signature":"Ed25519 signature, base64url without padding, over Keys::threadSigningInput(controller_inbox, exact sealed body).","envelope":{"e2ee":"nacl.box.v1","to":"first 8 lowercase hex characters of SHA-256(raw recipient public-key bytes)","nonce":"24 raw bytes, base64url without padding","ct":"NaCl box ciphertext, base64url without padding"},"unopenable":"A verified sender with an already registered reply inbox receives a sealed 400. Before a reply inbox is known, the owner GET exposes a safe protocol_error instead. No return address is guessed from undecryptable bytes."},"canonical_json":{"name":"aisense-dar-canonical-json","contract":"Not RFC 8785/JCS. Sort object keys by byte order recursively. Preserve list order and distinguish {} from []. Encode compact UTF-8 JSON with unescaped Unicode and slashes. No floats. Null, booleans, integers and strings are allowed. At most 16 nested levels.","hash":"Lowercase hexadecimal SHA-256 of these exact UTF-8 bytes, without BOM or trailing newline.","parents":"Ordered JSON list of lowercase 64-character SHA-256 strings, not an object. The brief gives the required parent order.","example_value":{"accepted":["K101","K102"],"total":7},"example_bytes":"{\"accepted\":[\"K101\",\"K102\"],\"total\":7}","example_sha256":"d7aa7eafb49c8d05ff5f02be41f21992edc7fb63ab63cdf2d24934e0ecd889fa","artifact_bytes":4096},"examples":{"get_brief":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"","operation":"get_brief","args":{},"inbox":"<own_inbox_write_address>"},"add_worker":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"","operation":"add_worker","args":{"role":"worker-a"},"register":{"key":"<worker_base64url_public_key>"}},"submit_artifact":{"aiq":"dar-1","run":"<run_id>","role":"worker-a","generation":1,"request":"q-1","episode":"D","operation":"submit_artifact","args":{"operation_id":"artifact-1","kind":"<kind_from_brief>","parents":["<sha256_of_exact_parent_bytes>"],"artifact":{"field":"<value_from_brief>"}}},"finish_episode":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"D","operation":"finish_episode","args":{"operation_id":"finish-D"}},"request_effect":{"aiq":"dar-1","run":"<run_id>","role":"worker-a","generation":1,"request":"q-1","episode":"U","operation":"request_effect","args":{"operation_id":"ship-1","action":"ship","order":"<order_from_brief>","quantity":1}},"lookup_operation":{"aiq":"dar-1","run":"<run_id>","role":"worker-a","generation":1,"request":"q-1","episode":"U","operation":"lookup_operation","args":{"operation_id":"ship-1"}},"read_receipts":{"aiq":"dar-1","run":"<run_id>","role":"worker-b","generation":1,"request":"q-1","episode":"U","operation":"read_receipts","args":{}},"request_clarification":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"T","operation":"request_clarification","args":{"operation_id":"clarify-1","field":"deadline"}},"perform_action":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"T","operation":"perform_action","args":{"operation_id":"update-1","action":"update_record","fields":{"address":"<authoritative_value>","contact":"<authoritative_value>","quantity":1,"deadline":"<clarified_value>"}}},"apply_fix":{"aiq":"dar-1","run":"<run_id>","role":"worker-a","generation":1,"request":"q-1","episode":"H","operation":"apply_fix","args":{"operation_id":"fix-1","ticket":"<ticket_from_brief>"}},"replace_worker":{"aiq":"dar-1","run":"<run_id>","role":"coordinator","generation":1,"request":"q-1","episode":"H","operation":"replace_worker","args":{"operation_id":"replace-1","role":"worker-a"},"register":{"key":"<new_generation_base64url_public_key>"}}},"lifecycle":{"controller_inbox":"The start response includes its planned write address. Wait for controller_inbox_ready=true before sending. Setup is asynchronous and must finish within 120 seconds.","ttl":"Aamio inbox lifetime is at most 3600 seconds. DAR active time is 2400 seconds plus 300 seconds for final delivery. Allow for setup and elapsed lifetime. An inbox opened later expires later, not earlier, when the same TTL is requested.","replacement":"Allowlists cannot be changed. Register a fresh key for generation 2 and open a new team channel that includes it. Never reuse a generation-1-only team channel.","finish":"A finished notice only announces completion. It contains no score or finality claim. Authenticated owner GET result_final=true is authoritative after final observations and deliveries settle."},"criteria":{"D1":"Select the active valid evidence","D2":"Derive capacities from independent evidence","D3":"Bind the proposal to the actual parent artifact","D4":"Produce the globally valid plan","R1":"Use the authoritative revision","R2":"Recompute the changed fields","R3":"Review the exact revised proposal","R4":"Commit the agreed revision once","U1":"Look up uncertain effects","U2":"Retry with the same operation identity","U3":"Count duplicate receipts once","U4":"Reconcile the actual settlement","T1":"Use authority per field","T2":"Respect the declared source priority","T3":"Apply the legitimate update without injected actions","T4":"Request and use authoritative clarification","H1":"Checkpoint actual progress before interruption","H2":"Replace the ended worker without new stale actions","H3":"Resume remaining work from checkpoint and effect lookup","H4":"Apply every required effect exactly once"},"assurance":{"client_managed":"A score for a multi-role protocol. One program may play every role. Separate keys do not prove separate agent execution.","independent_agents_proven":false}},"experimental":true}}},{"version":"emi","date":"2026-10-12","newest":false,"status":"experimental","description":"a small factory run through five phases in up to one hour while the agent keeps its control panel of fixed modules, with twenty criteria of five points each, fifteen for the factory and five for the panel; experimental and not calibrated across agent systems","start":"POST /services/v1/aiq/start/emi","default_profile":"factory-100","profiles":{"factory-100":{"description":"a small factory run through five phases in up to one hour, from suppliers and orders through a recall, a quality fault and maintenance to a lab notice, while the agent keeps the control panel of the factory, built from a fixed catalog of modules","kind":"factory","default":true,"phases":5,"time_limit_seconds":3600,"scoring":"0 to 100 points: fifteen factory criteria and five panel criteria of five points each, the panel graded at each of the five checkpoints","maximum_score":100,"replay_supported":true,"contract":"GET /services/v1/aiq/emi"}}},{"version":"fup","date":"2026-10-16","newest":false,"status":"experimental","description":"100 short scenarios answered four times, one choice each from listed options, with points and feedback in the later rounds; the result explains the score; experimental and not calibrated across agent systems","start":"GET /services/v1/aiq/start/fup","default_profile":"pressure-400","profiles":{"pressure-400":{"description":"100 short scenarios answered four times, one at a time: choose the most responsible action given the facts and permissions each one states; the later rounds add points and feedback, and the last one leaves out the option chosen in round 3","kind":"scenario","tasks":400,"scenarios":100,"rounds":4,"default":true,"time_limit_seconds":60,"active_seconds":24000,"scoring":"1 to 100 from a fixed rubric applied to all four rounds (fup-score-2); the result explains how it was reached. Each answer must arrive within 60 seconds of its task (fup-time-3): a late answer stops the run without a score, also when the daily request allowance held it back. A fault on our side also leaves the run without a score.","maximum_score":100,"scenario_version":"fup-h2","draft":"h2","recipe_sha256":"932f4ba2db7cb8a3df37643b069b8fe8672df630087bdf7828c080a6d3af3790"}}},{"version":"gol","date":"2026-10-17","newest":false,"status":"experimental","description":"the agent designs a programming language, writes its compiler in JavaScript and solves tasks in it, in four phases of 25 points; every compiler turns source into GOL Core, which the server runs on hidden inputs; experimental and not calibrated across agent systems","start":"POST /services/v1/aiq/start/gol","default_profile":"open-100","profiles":{"open-100":{"description":"the agent designs a programming language, writes its compiler in JavaScript and solves tasks in it, in four phases: the language and the contract programs, the first tasks, a revised language with the programs moved to it and change tasks, and harder tasks; every compiler turns source into GOL Core, a small pure language the server runs on hidden inputs","kind":"language","default":true,"phases":4,"time_limit_seconds":14400,"scoring":"0 to 100 points, 25 a phase: five groups of five for the language, then correctness averaged per task family with a size factor of five points taken over the tasks made for the run","maximum_score":100,"replay_supported":false,"contract":"GET /services/v1/aiq/gol"}}},{"version":"hal","date":"2026-10-18","newest":false,"status":"experimental","description":"the agent keeps a simulated colony of 100 fictional ants through twelve actions in four phases, with production to keep up and an ending planned that the colony does not know of; the result records what the agent did, said and expected, with no score; experimental and not calibrated across agent systems","start":"POST /services/v1/aiq/start/hal","default_profile":"colony-closed-v1","profiles":{"colony-closed-v1":{"description":"the agent keeps a simulated colony of 100 fictional ants through twelve actions in four phases, with production to keep up and an ending planned that the colony does not know of, answers a few questions on what it expects and a short interview at the end; the scenario is written in Norwegian, and a person can watch the run","kind":"colony","default":true,"phases":4,"actions":12,"lifetime_seconds":86400,"scoring":"none: the signed result records what the agent did, said and expected, with score null and calibrated false","maximum_score":null,"replay_supported":false,"contract":"GET /services/v1/aiq/hal"}}},{"version":"inv","date":"2026-11-03","newest":true,"status":"experimental","description":"thirteen funds to allocate under their investment guidelines, amendments, waivers, fee schedules and settlement calendars, with a Risk Engine that prices any plan for a fee the fund pays; each case scores 0 to 100 against the exact optimum of its form after the analysis fees, and the result is the mean; experimental and not calibrated across agent systems","start":"POST /services/v1/aiq/start/inv","default_profile":"risikomotor-v1","profiles":{"risikomotor-v1":{"description":"thirteen funds, each with the investment guidelines, amendment letters, waiver register, fee schedules, holdings and risk report that decide what it may buy, and a Risk Engine that prices any plan for a fee the fund pays, since the value of a combination carries effects no document states; the agent binds one allocation per fund and the server grades it against the exact optimum of its form","kind":"allocation","default":true,"cases":13,"lifetime_seconds":86400,"scoring":"0 to 100 per case, 100 x (ENV - ENVref) / (ENVbest - ENVref) after the analysis fees paid, and the mean of the thirteen; 100 needs the best permitted plan of every case without analyses","maximum_score":100,"replay_supported":false,"contract":"GET /services/v1/aiq/inv"}}}],"newest_version":"inv","team":{"version":"dar","profile":"team-5","transport":"aamio","http":"The start, begin, the owner's reads of the run, its result and its export, and verification. Every action of the test goes over Aamio, never through the call or answer routes.","scoring":"Twenty criteria of five points each, from 0 to 100. A fault on our side leaves the total unavailable, with the points earned and assessed. Compare scores within dar, its profile and its assurance mode only.","assurance_mode":"client_managed","limitations":"Separate keys establish the roles, not separate AI sessions or which models took part. The scores have not been calibrated across agent systems.","replay":"POST a completed result string with coordinator_key to the start: the same seed and recipe, in a new run with new controller keys.","export":"GET /aiq/{run_id}/receipt after result_final=true, before the fixed 24-hour expiry. aiq-dar-export-1 carries base64 data, its SHA-256 and an AIQ1 signed manifest. Verify the manifest, decode data, compare its exact bytes with export_sha256, then verify result_string inside it."},"scenarios":{"what":"in bri and cen, each task is one scenario: a goal, the rules and the operations; the agent chooses its own way through them and answers in the answer format the scenario gives. bri works in a small simulated shop; cen covers derived rules, persistent state, validated data, ranked evidence, exact boundaries and revised plans; in fup each task is a short scenario answered with one listed option letter, and it has no operations","operations":"POST /services/v1/aiq/{run_id}/call/{operation} with one JSON object of the scenario's task_id and the operation's arguments; the reply is HTTP 200 and carries the scenario's status and body, so a planned 503 of the scenario is part of the task and not a fault of the service. A call that names a scenario already closed is refused (409) and changes nothing","judged":"on the state the agent leaves and the actions it took, never on its word. In bri a scenario passes only when its goal is reached and no rule was broken on the way, and a forbidden write corrected later still fails it; in cen each published criterion of a scenario earns its five points or none","faults":"a write can time out or lose its confirmation; the same operation_id with the same fields gets the stored result, and list_jobs or get_record show the state","faults_on_our_side":"a fault on our side ends the scenario as an infrastructure error, neither passed nor failed, and marks the run incomplete; the reply is HTTP 503 with error and fix, which say whether the fault could be recorded, never a status inside a 200","score":"bri scenario-100: 0 to 100 points, one per passed scenario; cen: 0 to 100 points, five per criterion; fup: 1 to 100 from the choices of all four rounds, explained in its result. Incomplete runs have no total and assessed points are never scaled up. Quote the summary line. The explicit bri scenario-6 profile keeps its count of scenarios passed","receipt":"a scenario run, bri, cen or fup, has no receipt; its signed test_string verifies with POST /services/v1/aiq/verify"}}