Round r70-spot1

r70-spot1: Four-language round

Status · DESCRIPTIVEReviewed status is transcribed from rounds/STATUS.md. The original narrative and its qualifications remain below.

Source-reported narrativeRound README and results text are source material. Numeric values that are not tied to a claim ledger record remain source-reported.

Round source README · rounds/r70-spot1/README.md

An unregistered spot check ran two Round 70 tasks once in Rust, Elixir, Go, and TypeScript; six of eight cells were official full passes.

Round date: 2026-10-09 Publication badge: DESCRIPTIVE; KEPT FOR AUDIT Why not VALID: DESCRIPTIVE_ONLY, PRE_REGISTRATION_MISSING, RAW_EVIDENCE_PARTIAL, SMALL_SAMPLE — this was a small, unregistered spot check with incomplete evidence capture. Recomputation status: OBSERVED SOURCE ONLY

STATUS: DESCRIPTIVE

Why this status: This was a small, unregistered spot check with incomplete evidence capture. Its results describe these individual runs and do not estimate a language effect.

How and where it ran

Runs used two Hetzner CCX13 Linux hosts, one in Europe and one in the US. The cells ran on 2026-10-09, approximately 10:52Z–11:25Z UTC. The configured model was gpt-6-luna at max effort. The Round 70 runner used agent client version 0.161.0. Its launcher script stops a cell at 1,200 seconds (20 minutes); the Go task-7 launcher timestamps span 1,216 seconds from start to timeout/end. Round 70 task kit with restored prompts and prompt-hash preflight; the runner used agent CLI 0.161.0.

Task 5 ran on the Europe host and task 7 on the US host, once per language. The Go task-7 attempt is an official FAIL because it timed out (launcher rc 124); a separate grade of the tree retained after stopping passed 24/24 tests. The Elixir stage-probe labels were withdrawn because the public probe did not reproduce the grading environment. No replacements were recorded.

Assigned cells ran serially on each host. This spot check was not pre-registered.

Lifecycle counts

Phase Count
Planned 8
Started 8
Finished 8
Officially graded cells 8
ITT denominator 8

A cell is one attempt at one task in one language; intention-to-treat (ITT) keeps assigned attempts in the analysis and counts timeouts as failures.

Results

A full pass means the official task grade recorded a pass. Hidden suite test counts are shown as passed/total where they were captured; “0/—” means the tests did not run or no denominator is available. “After-stop tree grade” is reported separately from the official outcome. “Timeout” reports the runner time limit; “gate” reports the official full-pass result.

Cell outcomes

Cell Task ID Exact arm Model / effort Rep Full outcome Hidden suite tests passed / total After-stop tree grade Timeout / gate flag
spot1-eu-r70-5-elixir r70-5-elixir elixir gpt-6-luna / max 1 F 23/24 — timeout: no; gate: fail; rc 1
spot1-eu-r70-5-go r70-5-go go gpt-6-luna / max 1 P 24/24 — timeout: no; gate: pass; rc 0
spot1-eu-r70-5-rust r70-5-rust rust gpt-6-luna / max 1 P 24/24 — timeout: no; gate: pass; rc 0
spot1-eu-r70-5-ts-bun r70-5-ts-bun TypeScript/Bun gpt-6-luna / max 1 P 24/24 — timeout: no; gate: pass; rc 0
spot1-us-r70-7-elixir r70-7-elixir elixir gpt-6-luna / max 1 P 24/24 — timeout: no; gate: pass; rc 0
spot1-us-r70-7-go r70-7-go go gpt-6-luna / max 1 F 24/24 PASS (24/24) timeout: yes; gate: fail; rc 124
spot1-us-r70-7-rust r70-7-rust rust gpt-6-luna / max 1 P 24/24 — timeout: no; gate: pass; rc 0
spot1-us-r70-7-ts-bun r70-7-ts-bun TypeScript/Bun gpt-6-luna / max 1 P 24/24 — timeout: no; gate: pass; rc 0

Per-language summary

Rust had 2/2 official full passes, TypeScript (Bun) had 2/2, Elixir had 1/2, and Go had 1/2. Overall, six of eight attempts were official full passes. Go task 7 officially failed by timeout; its separate after-stop tree grade passed 24/24 tests.

Reproduction metadata

  • Kogen commit: not recorded in the surviving public source.
  • Harness commit: not recorded in the surviving public source.
  • Model and effort: gpt-6-luna at max.
  • Task IDs: r70-5-elixir, r70-5-go, r70-5-rust, r70-5-ts-bun, r70-7-elixir, r70-7-go, r70-7-rust, r70-7-ts-bun.
  • Historical command: exact invocation is not retained; the recorded setup used the Round 70 runner described above.
  • Raw records: per-cell Standard records; cell outcomes; filtered grade and audit extracts.

What's missing and why

  • circumstances.concurrent_cells_end (8 missing): Host-wide concurrent cell count was not captured Reason: Host-wide concurrent cell count was not captured (m66).
  • circumstances.concurrent_cells_start (8 missing): Host-wide concurrent cell count was not captured Reason: Host-wide concurrent cell count was not captured (m66).
  • circumstances.dispatcher_id (8 missing): Per-cell dispatcher revision is not present in the public round data Reason: Per-cell dispatcher revision is not present in the public round data (m72).
  • circumstances.incidents (8 missing): Per-cell incident receipt was not captured Reason: Per-cell incident receipt was not captured (m78).
  • circumstances.load1_end (8 missing): Host boundary load was not captured in the public round data Reason: Host boundary load was not captured in the public round data (m65).
  • circumstances.load1_start (8 missing): Host boundary load was not captured in the public round data Reason: Host boundary load was not captured in the public round data (m65).
  • circumstances.load_samples (8 missing): Timestamped host-wide load samples were not captured Reason: Timestamped host-wide load samples were not captured (m93).
  • circumstances.queue (8 missing): Queue receipt was not captured Reason: Queue receipt was not captured (m91).
  • cost.calculator_version (8 missing): Cost calculator version not recorded Reason: Cost calculator version not recorded (m61).
  • cost.price_table_version (8 missing): Price table version not recorded Reason: Price table version not recorded (m89).
  • cost.usd (8 missing): Per-cell cost and pricing inputs are not present in the public round data Reason: Per-cell cost and pricing inputs are not present in the public round data (m70).
  • environment.account_class (8 missing): Dated account-class receipt is not present in the public round data Reason: Dated account-class receipt is not present in the public round data (m62).
  • environment.network.allowlist_hosts (8 missing): Per-cell network allowlist not present in the public round data Reason: Per-cell network allowlist not present in the public round data (m79).
  • environment.toolchains.elixir (8 missing): Toolchain version not recorded Reason: Toolchain version not recorded (m95).
  • environment.toolchains.erlang (8 missing): Toolchain version not recorded Reason: Toolchain version not recorded (m95).
  • environment.toolchains.node (8 missing): Toolchain version not recorded Reason: Toolchain version not recorded (m95).
  • environment.toolchains.other_inventory (8 missing): Full per-cell toolchain inventory not recorded Reason: Full per-cell toolchain inventory not recorded (m64).
  • environment.toolchains.ruby (8 missing): Toolchain inventory not recorded Reason: Toolchain inventory not recorded (m94).
  • environment.toolchains.rust (8 missing): Toolchain version not recorded Reason: Toolchain version not recorded (m95).
  • grade.grader (8 missing): Per-cell grader software version is not in the public data Reason: Per-cell grader software version is not in the public data (m77).
  • grade.timestamp (8 missing): Per-cell official repair-grade timestamp is not present in the public data/cells.csv or sanitized lane release-state receipt Reason: Per-cell official repair-grade timestamp is not present in the public data/cells.csv or sanitized lane release-state receipt (m81).
  • sandbox.egress_allow (8 missing): Per-cell egress allowlist is not present in the public round data Reason: Per-cell egress allowlist is not present in the public round data (m75).
  • sandbox.profile (8 missing): Per-cell sandbox profile is not present in the public round data Reason: Per-cell sandbox profile is not present in the public round data (m87).
  • sandbox.profile_sha256 (8 missing): Per-cell sandbox profile fingerprint is not present in the public round data Reason: Per-cell sandbox profile fingerprint is not present in the public round data (m86).
  • setup.deps_source (8 missing): Per-cell dependency source is not recorded Reason: Per-cell dependency source is not recorded (m71).
  • setup.sandbox_mode (8 missing): Per-cell sandbox mode is not present in the public round data Reason: Per-cell sandbox mode is not present in the public round data (m85).
  • setup.sandbox_profile_sha256 (8 missing): Per-cell sandbox fingerprint is not present in the public round data Reason: Per-cell sandbox fingerprint is not present in the public round data (m84).
  • timestamps.attempts (8 missing): Attempt-boundary receipts are not in the public round data Reason: Attempt-boundary receipts are not in the public round data (m60).
  • timestamps.phases.develop.end_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.develop.start_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.gate.end_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.gate.start_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.grade.end_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.grade.start_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.plan.end_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.plan.start_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.review.end_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.review.start_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.setup.end_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.setup.start_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.shape.end_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timestamps.phases.shape.start_utc (8 missing): Phase timestamp not present in the public per-cell data Reason: Phase timestamp not present in the public per-cell data (m88).
  • timing.phases_s.develop (8 missing): develop phase timing was not separately recorded in the public round data Reason: develop phase timing was not separately recorded in the public round data (m96).
  • timing.phases_s.grade (8 missing): grade phase timing was not separately recorded in the public round data Reason: grade phase timing was not separately recorded in the public round data (m98).
  • timing.phases_s.setup (8 missing): setup phase timing was not separately recorded in the public round data Reason: setup phase timing was not separately recorded in the public round data (m99).
  • tokens.phases.develop.cached_input (8 missing): develop phase token counter was not recorded separately Reason: develop phase token counter was not recorded separately (m97).
  • tokens.phases.develop.input (8 missing): develop phase token counter was not recorded separately Reason: develop phase token counter was not recorded separately (m97).
  • tokens.phases.develop.output (8 missing): develop phase token counter was not recorded separately Reason: develop phase token counter was not recorded separately (m97).
  • tokens.phases.develop.reasoning (8 missing): develop phase token counter was not recorded separately Reason: develop phase token counter was not recorded separately (m97).
  • tokens.total.cached_input (1 missing): develop phase token counter was not recorded separately Reason: develop phase token counter was not recorded separately (m97).
  • tokens.total.input (1 missing): develop phase token counter was not recorded separately Reason: develop phase token counter was not recorded separately (m97).
  • tokens.total.output (1 missing): develop phase token counter was not recorded separately Reason: develop phase token counter was not recorded separately (m97).
  • tokens.total.reasoning (1 missing): develop phase token counter was not recorded separately Reason: develop phase token counter was not recorded separately (m97).
  • tools.grader (8 missing): Per-cell grader software version is not in the public data Reason: Per-cell grader software version is not in the public data (m77).
  • tools.runner (8 missing): Per-cell grader software version is not in the public data Reason: Per-cell grader software version is not in the public data (m77).
  • tools.toolchains.elixir (8 missing): Toolchain version not recorded Reason: Toolchain version not recorded (m95).
  • tools.toolchains.erlang (8 missing): Toolchain version not recorded Reason: Toolchain version not recorded (m95).
  • tools.toolchains.node (8 missing): Toolchain version not recorded Reason: Toolchain version not recorded (m95).
  • tools.toolchains.other_inventory (8 missing): Full per-cell toolchain inventory not recorded Reason: Full per-cell toolchain inventory not recorded (m64).
  • tools.toolchains.ruby (8 missing): Toolchain inventory not recorded Reason: Toolchain inventory not recorded (m94).
  • tools.toolchains.rust (8 missing): Toolchain version not recorded Reason: Toolchain version not recorded (m95).

Records and methods

The per-cell Standard records are in records.jsonl. The complete result table and launcher details are in RESULTS.md; filtered grade fields, audit counts, and launcher extracts are in pulled-all. See the measurement contract and the review record.

Round results · rounds/r70-spot1/RESULTS.md

DESCRIPTIVE SPOT CHECK; not a registered comparison. Two tasks (r70-5 in Europe and r70-7 in the US) were run once in each of four languages. The retained launcher script sets a 1,200-second (20-minute) limit; the launcher log records 1,216 seconds between the Go task-7 start and timeout/end timestamps. Results are per cell.

The Go r70-7 attempt is an official FAIL because the launcher timed out (rc 124). A separate grade of the tree retained after stopping is PASS (24/24 tests); this does not change the official timeout outcome. The retained launcher script sets the cutoff at 1,200 seconds; its log timestamps span 1,216 seconds from start to timeout/end. Firewall rejection deltas were constant within each host: 30 per Europe cell and 72 per US cell.

Cell Task Exact arm label Official outcome Tests Failures + errors Tests ran After-stop tree grade Launcher rc Wall s Firewall rejects
spot1-eu-r70-5-elixir r70-5-elixir elixir FAIL 24 1 yes — 1 410 30
spot1-eu-r70-5-go r70-5-go go PASS 24 0 yes — 0 230 30
spot1-eu-r70-5-rust r70-5-rust rust PASS 24 0 yes — 0 210 30
spot1-eu-r70-5-ts-bun r70-5-ts-bun ts-bun PASS 24 0 yes — 0 231 30
spot1-us-r70-7-elixir r70-7-elixir elixir PASS 24 0 yes — 0 410 72
spot1-us-r70-7-go r70-7-go go FAIL (timeout) 24 0 yes PASS (24/24) 124 1216 72
spot1-us-r70-7-rust r70-7-rust rust PASS 24 0 yes — 0 180 72
spot1-us-r70-7-ts-bun r70-7-ts-bun ts-bun PASS 24 0 yes — 0 160 72

The corrected audit produced no flagged cells. Per-cell audit details are in `audit-corrected.jsonl`; the retained audit outputs are in `out/`. A stage-probe attempt to label Elixir failures was withdrawn because the public Elixir environment did not reproduce the grading environment. No Elixir stage labels are used here.

Captured run records

Standard-schema records assigned to this round in the indexed snapshot.

No per-round file is listed. Check the run-record index for the complete partition list.

Reproduction inventory

Published analysis replay and full experiment replay are reported separately.

Replay type
unavailable
Script
No per-round analysis script is published.
Inputs
No identified public input bundle.
Full execution replay
Unavailable
Scope
Only filtered grade and manifest fields, selected launcher lines, and corrected audit counts are available; no complete delivery bundle or execution replay is published.

View the complete per-round reproduction inventory · Repository reproduction guide