✔ continuation reads advance through historical pages without a false pause while repeated current-Run pages still pause (1.8342ms) ✔ a new tool page replaces old latest-feedback space and still obeys its escaped protocol budget (0.6263ms) ✔ budgeted artifact pages count escaped protocol bytes and reconstruct original multilingual JSON without losing characters (9.2333ms) ✔ known artifact coverage detects semantic repeats despite changing limits, and advancing pages or direct evidence restore progress (0.3434ms) ✔ actual large artifact reads are present in the next bounded HTTP request and next_offset progresses without redoing tools (7168.2766ms) ✔ a real repeated artifact loop pauses the same Task after recorded results, survives reload and only resumes explicitly (7126.1239ms) ✔ actual HTTP continuation after reload sends distinct historically covered pages without false no_progress or replay (7218.8157ms) ✔ native seed framing reserves capacity before optional history without changing the selected window (2024.8494ms) ✔ direct code evidence remains selected while recent historical-index browsing is archived within the complete request budget (3087.7985ms) ✔ one long user turn becomes bounded complete groups with exact UTF-8 history and recent JSON arguments (5799.3675ms) ✔ tool definitions, exact system bytes, output reserve and remaining Run/Task budget are accounted before requests (1093.7909ms) ✔ oversized required user instructions pause intact and can be explicitly resumed after increasing the window (1965.754ms) ✔ all task instructions, constraints and acceptance remain in the runtime system instead of an eight-item or byte prefix (1585.0679ms) ✖ a newest manageable multi-tool group stays exact even beyond the working target; oversized earlier groups are archived (1171.4455ms) ✔ invalid orphan result is rejected before archive or network preparation (787.392ms) ✔ invalid missing result is rejected before archive or network preparation (777.1528ms) ✔ invalid wrong result is rejected before archive or network preparation (773.6971ms) ✔ invalid duplicate call is rejected before archive or network preparation (750.3214ms) ✔ a group larger than the artifact limit uses verified Unicode-safe chunks without replacing tool JSON (2390.9915ms) ✔ changed or missing archived evidence stops reuse and is never silently recreated (1109.0773ms) ✔ transaction failure prevents publishing a plan and any corresponding model request (957.0372ms) ✔ the Provider rejects body mutation after a sealed plan before calling HTTP (1192.0419ms) ✔ HTTP failure and explicit same-Task resume reuse exact completed results and request hashes without repeating a command (9590.2848ms) ✔ a real process exit after plan persistence marks interruption and never issues an automatic model request (1583.4676ms) ✔ the latest oversized session result reaches the next request together with already read source modules (4240.8591ms) ✔ every call in a large latest group keeps its actual outcome, including denied and unknown results (2334.4132ms) ✔ a small complete latest tool group is sent once and needs no duplicate feedback view (1822.7439ms) ✔ real HTTP repeated state inspection pauses after three identical engineering outcomes and preserves tool pairing (4930.5822ms) ✔ small code and actual failed command stay visible when their original multi-tool group cannot fit (3487.7511ms) ✔ mixed group retains six distinct small file observations with exact multilingual text and bounded request bytes (3612.0072ms) ✔ working observations are bound to the current Task and stop treating a pre-application file revision as current evidence (3276.0903ms) ✔ a sealed observation rejects changed or missing original result artifacts instead of recreating evidence (2023.1842ms) ✔ a real passed-check observation is excluded after a source edit invalidates the check, while its historical sent evidence remains valid (7235.6425ms) ✔ budgeted file pages reconstruct UTF-8 text including long single lines and surrogate boundaries (4.203ms) ✔ actual HTTP multi-tool reads survive an oversized runtime snapshot and the sealed evidence survives reopening (13892.5968ms) ✔ actual HTTP file pagination advances over a long file with each returned page visible in the next request (4120.3793ms) ✔ later mixed history pages cannot displace the actual module sources and complete requirements (4872.1759ms) ✔ a later narrow page of the same file revision cannot evict the wider actual observation (3008.7191ms) ✔ six small source files stay complete across metadata turns even when mandatory evidence exceeds the soft target (2781.0973ms) ✔ historical read-artifact navigation points to the original artifact and next page, never its result wrapper (1151.7872ms) ✔ a 1.6.1 full-provenance sent view remains verifiable after introducing the compact view (1170.9574ms) ✔ actual HTTP requests keep four observed modules after a later large history read and preserve all source bytes (23827.7338ms) ℹ tests 42 ℹ suites 0 ℹ pass 41 ℹ fail 1 ℹ cancelled 0 ℹ skipped 0 ℹ todo 0 ℹ duration_ms 74660.6056 ✖ failing tests: test at tests\request-plan.test.mjs:127:1 ✖ a newest manageable multi-tool group stays exact even beyond the working target; oversized earlier groups are archived (1171.4455ms) AssertionError [ERR_ASSERTION]: The expression evaluated to a falsy value: assert.ok(plan.record.estimatedInputTokens > plan.record.budget.targetInputTokens) at TestContext. (file:///D:/Ai-project/REA%E9%80%86%E5%90%91/Agent-Workbench/tests/request-plan.test.mjs:136:10) at Test.runInAsyncScope (node:async_hooks:227:14) at Test.run (node:internal/test_runner/test:1402:25) at Test.processPendingSubtests (node:internal/test_runner/test:974:18) at Test.postRun (node:internal/test_runner/test:1542:19) at Test.run (node:internal/test_runner/test:1467:12) at async Test.processPendingSubtests (node:internal/test_runner/test:974:7) { generatedMessage: true, code: 'ERR_ASSERTION', actual: false, expected: true, operator: '==', diff: 'simple' }