✔ Messages builds native system/tool history, bounds images, and seals a distinct mixed-pool variant (6.5484ms) ✔ Messages URL and native parameter contracts preserve other protocol paths and hard output budget (0.5985ms) ✔ native usage includes both cache buckets once, retains raw and rejects invalid counters (0.1786ms) ✔ real HTTP Messages JSON executes one tool, retains opaque history, actual URL and cache accounting (4313.6667ms) ✔ real HTTP Messages SSE executes one tool, retains opaque history, actual URL and cache accounting (3447.717ms) ✔ Messages api-key authentication is explicit; mismatched model remains blocked with original usage (6.1309ms) ✔ Messages bearer authentication is explicit; mismatched model remains blocked with original usage (4.5811ms) ✔ Messages redirect is not followed, either credential stays at the original server (3.6997ms) ✔ strict native SSE reads the nested start model and retains mismatch source without retries (3.0908ms) ✔ a closed native tool block cannot execute before message_stop, then executes exactly once (3213.8544ms) ✔ mixed Chat/Messages pool seals both variants and preserves one completed tool on failover (4246.1155ms) ✔ native JSON rejects incomplete stops, duplicate tools and hosted content (0.353ms) ▶ strict SSE validates all blocks, terminal, event names, UTF8 and tool JSON before dispatch ✔ reject native stream 0 (1.0888ms) ✔ reject native stream 1 (0.2255ms) ✔ reject native stream 2 (0.2871ms) ✔ reject native stream 3 (0.1879ms) ✔ reject native stream 4 (0.1582ms) ✔ reject native stream 5 (0.1619ms) ✔ reject native stream 6 (0.1227ms) ✔ reject native stream 7 (0.1105ms) ✔ reject native stream 8 (0.1843ms) ✔ strict SSE validates all blocks, terminal, event names, UTF8 and tool JSON before dispatch (3.1428ms) ✔ upstream SSE error retains original type/request ID and initial usage, never retries (4.7539ms) ✔ stop cancels the owned stream and retains partial consumption without running tool or replaying (61.3452ms) ✔ saved Messages import preserves auth choice, rejects endpoint duplication, and keeps source file untouched (958.0128ms) ✔ input that fits the engine window but exceeds remaining credit diagnoses the credit limit (0.5363ms) ✔ window overflow and historical pauses retain distinct and honest diagnoses (0.0809ms) ✔ R observed reservation shortage keeps the window-fit diagnosis separate from Run/Task credit (0.093ms) ✔ a budget pause persists concrete text/tool/history/image breakdown and never dispatches an oversized request (4056.5419ms) ✔ explicit frozen preview resumes the same Task in a new Run, preserves completed tools and does not change global config or selected model (4375.4502ms) ✔ insufficient budgets, silently clamped capacity and arbitrary config fields reject without a physical request or mutating task budget (2918.8302ms) ✔ changed session, changed registry, expiry and altered frozen preview all require a new explicit preview (3588.3646ms) ✔ preview preserves unknown invocation status and increasing Task allowance is committed only with confirmed new Run (5129.1533ms) ✔ saved engine budgets verify all Chat/Responses route variants and reject a mismatched output allowance (1075.63ms) ✔ owned CC Switch SQLite imports the selected TOML Responses wire_api and leaves unknown clients untouched (1455.6215ms) ✔ Responses executes one original tool through Runtime, carries encrypted Items, reports raw usage and seals exact wire bytes (3058.5134ms) ✔ mixed-protocol pool keeps both actual variants and changes route on 429 without re-executing completed tools (3965.986ms) ▶ Responses incomplete, failed, missing call identity and unsupported hosted outputs never dispatch a tool ✔ Reject {"body":{"id":"resp_owned","status":"incomplete","output":[{"type":"function_call","id":"fc_partial" (2033.5475ms) ✔ Reject {"body":{"id":"resp_owned","status":"failed","output":[{"type":"function_call","id":"fc_failed","cal (1919.3653ms) ✔ Reject {"body":{"id":"resp_owned","status":"completed","output":[{"type":"function_call","id":"fc_bad","cal (2027.8655ms) ✔ Reject {"body":{"id":"resp_owned","status":"completed","output":[{"type":"computer_call","id":"not-supporte (1631.6839ms) ✔ Responses incomplete, failed, missing call identity and unsupported hosted outputs never dispatch a tool (7613.4068ms) ▶ invalid JSON in a nonstream Chat or Responses tool group prevents every tool and pool replay ✔ chat-completions (1451.0538ms) ✔ responses (1388.3246ms) ✔ invalid JSON in a nonstream Chat or Responses tool group prevents every tool and pool replay (2839.7747ms) ✔ wire adapter preserves tool IDs but never passes Responses-private Items to Chat or another endpoint (0.2109ms) ✔ connection JSON import accepts explicit Responses protocol and continues to reject unknown protocols without reading client sessions (618.8068ms) ✔ reservation uses actual known usage from the same route, never lowers estimates or rewrites original receipts (1.5861ms) ✔ reservation verification binds original Task/Run consumption and rejects changed usage or foreign provenance (0.4653ms) ✔ known high input calibrates the next request and stops before dispatch while preserving the completed read (1951.0713ms) ✔ one reported overrun stops legacy requests and keeps actual use and excess (1509.2561ms) ✔ one reported overrun stops pool requests and keeps actual use and excess (2153.2021ms) ✔ one reported overrun stops connection requests and keeps actual use and excess (2059.8224ms) ✔ reported usage below the limit prevents an unaffordable legacy retry before another dispatch (1713.5456ms) ✔ reported usage below the limit prevents an unaffordable pool retry before another dispatch (2060.3406ms) ✔ native bridge seals calibrated refusal with original consumption and preserves versioned request verification (12821.7929ms) ✔ actual HTTP requests reuse only this project source-checked memory as redacted data and preserve sent evidence after archive/restart (1701.3743ms) ✔ actual HTTP request excludes archived, timed-out and foreign-project memories (2694.8128ms) ✔ actual next HTTP request removes expired memory after a tool observes a source change without replaying that tool (2237.4216ms) ✔ incomplete inventory and excluded-file SHA changes cannot enter actual HTTP memory context or inspect_session snapshots (2585.9887ms) ✔ corrupt original artifact expires its memory and cannot reach an HTTP model request (1997.7985ms) ✔ memory JSON is bounded in escaped UTF-8 bytes and entry count, skips oversized entries and retains complete multilingual values (4106.7101ms) ✔ optional memory yields to the complete goal and original code evidence under the whole-request budget (1376.0617ms) ✔ changed sealed memory artifacts or binding fields fail verification while the original request remains immutable (2487.6342ms) ✔ a malformed project declaration omits optional memory while the repair Task still reaches the owned HTTP service (2576.1328ms) ✔ an expiration persistence failure prevents dispatch and is not hidden as an optional scan failure (1944.4406ms) ✔ continuation reads advance through historical pages without a false pause while repeated current-Run pages still pause (1.8851ms) ✔ a new tool page replaces old latest-feedback space and still obeys its escaped protocol budget (0.4571ms) ✔ budgeted artifact pages count escaped protocol bytes and reconstruct original multilingual JSON without losing characters (9.0225ms) ✔ known artifact coverage detects semantic repeats despite changing limits, and advancing pages or direct evidence restore progress (0.5303ms) ✔ actual large artifact reads are present in the next bounded HTTP request and next_offset progresses without redoing tools (4568.8846ms) ✔ a real repeated artifact loop pauses the same Task after recorded results, survives reload and only resumes explicitly (9154.0295ms) ✔ actual HTTP continuation after reload sends distinct historically covered pages without false no_progress or replay (9290.3679ms) ✔ continuations distinguish unread source, later original line ranges and already returned result wrappers (1391.3603ms) ✔ pure-file batch returns unused small-page space while mixed and historical allocation stays conservative (0.3365ms) ✔ observed upstream high input pauses before a new request, preserves reported usage and distinguishes the limiting ledger (1119.5178ms) ✔ chat-completions: complete actual Unicode/CRLF pages and small source survive the next bounded wire request (4489.4919ms) ✔ responses: complete actual Unicode/CRLF pages and small source survive the next bounded wire request (5136.0782ms) ✔ anthropic-messages: complete actual Unicode/CRLF pages and small source survive the next bounded wire request (4040.5572ms) ✔ a real two-reply failure replay reaches request three at 16K with complete goal, original sources and sealed tool evidence (10517.5929ms) ✔ request deduplication retains long original intent, extra requirements, constraints and acceptance without changing the stored contract (2958.3216ms) ✔ only the routine context avoids plan echoes; diagnostic API and requests without an included plan keep original observations (826.171ms) ✔ small code and actual failed command stay visible when their original multi-tool group cannot fit (3228.0359ms) ✔ mixed group retains six distinct small file observations with exact multilingual text and bounded request bytes (3286.0909ms) ✔ working observations are bound to the current Task and stop treating a pre-application file revision as current evidence (2421.2836ms) ✔ a sealed observation rejects changed or missing original result artifacts instead of recreating evidence (1558.8453ms) ✔ a real passed-check observation is excluded after a source edit invalidates the check, while its historical sent evidence remains valid (5042.6206ms) ✔ budgeted file pages reconstruct UTF-8 text including long single lines and surrogate boundaries (4.7148ms) ✔ actual HTTP multi-tool reads survive an oversized runtime snapshot and the sealed evidence survives reopening (11973.8585ms) ✔ actual HTTP file pagination advances over a long file with each returned page visible in the next request (6994.9477ms) ✔ later mixed history pages cannot displace the actual module sources and complete requirements (6745.4624ms) ✔ a later narrow page of the same file revision cannot evict the wider actual observation (3898.4259ms) ✔ six small source files stay complete across metadata turns even when mandatory evidence exceeds the soft target (3601.4413ms) ✔ historical read-artifact navigation points to the original artifact and next page, never its result wrapper (966.1327ms) ✔ a 1.6.1 full-provenance sent view remains verifiable after introducing the compact view (1104.2318ms) ✔ actual HTTP requests keep four observed modules after a later large history read and preserve all source bytes (9402.6698ms) ℹ tests 96 ℹ suites 0 ℹ pass 96 ℹ fail 0 ℹ cancelled 0 ℹ skipped 0 ℹ todo 0 ℹ duration_ms 137955.7445