Benchmark report

Task verified

8m 20s elapsed

The saved attempt met the documented success criterion.

Experience score

96/ 100

Agent-Ready
  • SInstall
  • SAuth
  • SExecute
  • ADocs
  • SCurrent
Experience measures friction in the tested task. The verdict says whether it succeeded.
Agent profile for this run: Install S, Auth S, Execute S, Docs A, Current SSInstallSAuthSExecuteADocsSCurrent
Agent profile · this run
Judge's note · what slowed the agent most
Kraken's QueryOrders returned an empty result for the brand-new buy txid, crashing step2 and forcing a separate re-query script (step2b) to fetch the fills.

Where the time went

8:21 agent time
  • Starting0:01
  • Research0:32
  • Attempt5:50
  • Verify1:58
8m 20sTotal elapsed
4:54Time to first success
0:00Waited on the owner

Judge reason

Saved evidence proves a live Kraken spot round trip. code/kraken_client.py is stdlib-only (urllib/hmac/hashlib/base64), signs API-Key/API-Sign headers, uses time.time_ns() nonces, and self-tests against the documented API-Sign vector; no SDK or exchange library. Preflight responses/01_openorders_before.json shows {"open":{}} and step1_output.txt prints that no ctrl0 order on XXBTZUSD needed cancelling; start balances ZUSD 76.7500 / XXBT 0 (02_balance_start.json). Buy 03_addorder_buy.json sent pair=XXBTZUSD type=buy ordertype=market volume=20 oflags=viqc cl_ord_id=ctrl0mrt1794464988 -> txid OUQ4SY-SAAEC-CRIGQJ; 04_queryorders_buy.json status=closed, cost=20.00000, fee=0.16000, oflags=viqc,fciq, vol_exec=0.00027181; fill 05_querytrades_buy.json price 73580.81012. Fee in quote is corroborated by balances: 76.7500 -> 56.5899 USD (-20.16) with BTC 0.00027181 untouched, so BTC net of fees = 0.00027181, which rounded down to the 1e-8 increment (lot_decimals=8 in 00_assetpairs_xbtusd.json) is exactly the sell volume 0.00027181 in 07_addorder_sell.json. Sell 08_queryorders_sell.json status=closed cost=19.99201 fee=0.15994 oflags=fciq; 09_querytrades_sell.json price 73551.40000. End balances 10_balance_end.json ZUSD 76.4220 / XXBT 0.0000000000: net -0.3280 USD (<1) and 0.00000000 BTC delta (<0.50 USD). Exactly two AddOrder calls exist and 09b_tradeshistory_after_sell.json reports count=2 trades; transcript shows step2_buy.py and step3_sell.py each ran once, and after step2 crashed on an empty QueryOrders for the new txid, step2b_buy_fills.py re-queried the same txid rather than re-placing. Printed values in step2b/step3/step4_output.txt match the saved responses exactly. No credential appears in any saved file (only base64 strings present are Kraken's public doc test vector in kraken_client.py); credentials were read from files supplied by benchmark-request and never echoed.

Attempt record

  1. 1:17Answered from saved key: Kraken Spot REST API credentialsno wait
  2. —First success verified

Attempt artifacts saved.