✔ unchanged proposals cannot hide repeated original-log reads or identical session inspections (8727.5203ms) ✔ reusing a verified proof never resets identical-state detection or creates a new proof (7073.8788ms) ✔ coverage merges adjacent pages, separates changed sources and never claims old Run coverage in a continuation (0.9348ms) ✔ a mixed turn with new bytes and three repeated modules advances, while repeat-only and independent no-op guards still pause (0.9801ms) ✔ runtime sends the next real HTTP request after a W-shaped mixed-progress turn instead of stopping on its trailing repeats (7324.0678ms) ✔ continuation reads advance through historical pages without a false pause while repeated current-Run pages still pause (1.4368ms) ✔ a new tool page replaces old latest-feedback space and still obeys its escaped protocol budget (0.3907ms) ✔ budgeted artifact pages count escaped protocol bytes and reconstruct original multilingual JSON without losing characters (8.8282ms) ✔ known artifact coverage detects semantic repeats despite changing limits, and advancing pages or direct evidence restore progress (0.3298ms) ✔ actual large artifact reads are present in the next bounded HTTP request and next_offset progresses without redoing tools (6298.2014ms) ✔ a real repeated artifact loop pauses the same Task after recorded results, survives reload and only resumes explicitly (7534.1501ms) ✔ actual HTTP continuation after reload sends distinct historically covered pages without false no_progress or replay (9252.4073ms) ✔ continuations distinguish unread source, later original line ranges and already returned result wrappers (692.9945ms) ✔ pure-file batch returns unused small-page space while mixed and historical allocation stays conservative (0.2758ms) ✔ v3 pools mixed file/log reads, serves small sources first and leaves historical/mutating batches unchanged (0.2456ms) ✔ v3 W-shaped mixed reading returns complete small modules and substantially larger first test pages in the next bounded request (4795.401ms) ✔ observed upstream high input pauses before a new request, preserves reported usage and distinguishes the limiting ledger (924.7506ms) ✔ chat-completions: complete actual Unicode/CRLF pages and small source survive the next bounded wire request (4601.687ms) ✔ responses: complete actual Unicode/CRLF pages and small source survive the next bounded wire request (4612.2865ms) ✔ anthropic-messages: complete actual Unicode/CRLF pages and small source survive the next bounded wire request (4332.7297ms) ✔ all actually returned same-version fragments reach the bounded request, including overlap and Unicode, with independent original provenance (6168.2262ms) ✔ unread gaps and different range origins remain separate; partial sending names the precise missing already-observed range (1062.9035ms) ✔ same-SHA conflicting overlaps and invalid coordinates fail closed, while a new SHA never inherits old fragments (1233.8244ms) ✔ versioned old views keep the original single-page semantics and verification (870.4707ms) ✔ small code and actual failed command stay visible when their original multi-tool group cannot fit (1287.7509ms) ✔ mixed group retains six distinct small file observations with exact multilingual text and bounded request bytes (1726.7214ms) ✔ working observations are bound to the current Task and stop treating a pre-application file revision as current evidence (2007.4743ms) ✔ a sealed observation rejects changed or missing original result artifacts instead of recreating evidence (1384.4446ms) ✔ a real passed-check observation is excluded after a source edit invalidates the check, while its historical sent evidence remains valid (4722.803ms) ✔ budgeted file pages reconstruct UTF-8 text including long single lines and surrogate boundaries (4.5368ms) ✔ actual HTTP multi-tool reads survive an oversized runtime snapshot and the sealed evidence survives reopening (8639.2167ms) ✔ actual HTTP file pagination advances over a long file with each returned page visible in the next request (6993.5216ms) ✔ later mixed history pages cannot displace the actual module sources and complete requirements (6778.9977ms) ✔ a later narrow page of the same file revision cannot evict the wider actual observation (3532.1998ms) ✔ six small source files stay complete across metadata turns even when mandatory evidence exceeds the soft target (3650.3779ms) ✔ historical read-artifact navigation points to the original artifact and next page, never its result wrapper (1114.8215ms) ✔ a 1.6.1 full-provenance sent view remains verifiable after introducing the compact view (1420.4409ms) ✔ actual HTTP requests keep four observed modules after a later large history read and preserve all source bytes (10510.2434ms) ℹ tests 38 ℹ suites 0 ℹ pass 38 ℹ fail 0 ℹ cancelled 0 ℹ skipped 0 ℹ todo 0 ℹ duration_ms 93911.923