{ "name": "FrontierHarness Eval v1", "status": "frozen", "model": "Kimi K3", "model_provider": "Fireworks", "task_count": 30, "harness_count": 9, "configuration_count": 12, "evaluation_count": 360, "canonical_selection": "first valid attempt 1", "excluded_labels": [ "warm", "smoke" ], "task_sources": { "terminal_bench": { "dataset": "terminal-bench/terminal-bench-2-1", "count": 21 }, "deep_swe": { "repository": "datacurve-ai/deep-swe", "version": "v1.1", "reproduction_ref": "435ee89ec2f2e2289f33b0da4f992f0b7b7266b9", "reproduction_note": "v1.1 is the published dataset label, not an upstream Git tag. The reproduction ref uses separate-verifier images that differ from the frozen public task metadata; validate comparability with a control run.", "count": 9 } }, "harness_configurations": [ "claude-code", "codex", "dsh-creator", "dsh-minimal", "dsh-ptc", "dsh-standard", "exo", "hermes", "kimi-code", "oh-my-pi", "opencode", "pi-responses" ], "results": "results/eval-data.json", "interactive_report": "https://frontierharness.org", "task_ids": [ "datacurve/anko-typed-variable-bindings", "datacurve/arktype-json-schema-refs-dependencies", "datacurve/expr-try-catch-errors", "datacurve/fastapi-deprecation-response-headers", "datacurve/httpx-multipart-response-parsing", "datacurve/katex-multicolumn-array-spans", "datacurve/meriyah-explicit-resource-declarations", "datacurve/python-statemachine-state-data-scoping", "datacurve/scc-bounded-memory-spilling", "terminal-bench/build-cython-ext", "terminal-bench/chess-best-move", "terminal-bench/code-from-image", "terminal-bench/constraints-scheduling", "terminal-bench/db-wal-recovery", "terminal-bench/dna-insert", "terminal-bench/extract-elf", "terminal-bench/gcode-to-text", "terminal-bench/git-leak-recovery", "terminal-bench/kv-store-grpc", "terminal-bench/largest-eigenval", "terminal-bench/log-summary-date-ranges", "terminal-bench/merge-diff-arc-agi-task", "terminal-bench/modernize-scientific-stack", "terminal-bench/multi-source-data-merger", "terminal-bench/openssl-selfsigned-cert", "terminal-bench/polyglot-c-py", "terminal-bench/regex-log", "terminal-bench/sanitize-git-repo", "terminal-bench/sqlite-db-truncate", "terminal-bench/vulnerable-secret" ] }