self-bench 0.3.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/LICENSE +21 -0
- package/README.md +256 -0
- package/dist/activities.d.ts +41 -0
- package/dist/activities.d.ts.map +1 -0
- package/dist/activities.js +823 -0
- package/dist/activities.js.map +1 -0
- package/dist/agent-smoke-main.d.ts +3 -0
- package/dist/agent-smoke-main.d.ts.map +1 -0
- package/dist/agent-smoke-main.js +54 -0
- package/dist/agent-smoke-main.js.map +1 -0
- package/dist/agent-smoke.d.ts +17 -0
- package/dist/agent-smoke.d.ts.map +1 -0
- package/dist/agent-smoke.js +98 -0
- package/dist/agent-smoke.js.map +1 -0
- package/dist/api-main.d.ts +2 -0
- package/dist/api-main.d.ts.map +1 -0
- package/dist/api-main.js +11 -0
- package/dist/api-main.js.map +1 -0
- package/dist/api.d.ts +3 -0
- package/dist/api.d.ts.map +1 -0
- package/dist/api.js +212 -0
- package/dist/api.js.map +1 -0
- package/dist/artifacts.d.ts +32 -0
- package/dist/artifacts.d.ts.map +1 -0
- package/dist/artifacts.js +311 -0
- package/dist/artifacts.js.map +1 -0
- package/dist/audit.d.ts +12 -0
- package/dist/audit.d.ts.map +1 -0
- package/dist/audit.js +69 -0
- package/dist/audit.js.map +1 -0
- package/dist/cli.d.ts +3 -0
- package/dist/cli.d.ts.map +1 -0
- package/dist/cli.js +297 -0
- package/dist/cli.js.map +1 -0
- package/dist/codex-review.d.ts +40 -0
- package/dist/codex-review.d.ts.map +1 -0
- package/dist/codex-review.js +189 -0
- package/dist/codex-review.js.map +1 -0
- package/dist/config.d.ts +35 -0
- package/dist/config.d.ts.map +1 -0
- package/dist/config.js +78 -0
- package/dist/config.js.map +1 -0
- package/dist/contracts.d.ts +223 -0
- package/dist/contracts.d.ts.map +1 -0
- package/dist/contracts.js +103 -0
- package/dist/contracts.js.map +1 -0
- package/dist/coupling.d.ts +45 -0
- package/dist/coupling.d.ts.map +1 -0
- package/dist/coupling.js +159 -0
- package/dist/coupling.js.map +1 -0
- package/dist/docker-executor.d.ts +11 -0
- package/dist/docker-executor.d.ts.map +1 -0
- package/dist/docker-executor.js +103 -0
- package/dist/docker-executor.js.map +1 -0
- package/dist/eval-main.d.ts +3 -0
- package/dist/eval-main.d.ts.map +1 -0
- package/dist/eval-main.js +80 -0
- package/dist/eval-main.js.map +1 -0
- package/dist/evaluate.d.ts +24 -0
- package/dist/evaluate.d.ts.map +1 -0
- package/dist/evaluate.js +208 -0
- package/dist/evaluate.js.map +1 -0
- package/dist/extensions/authoring.d.ts +3 -0
- package/dist/extensions/authoring.d.ts.map +1 -0
- package/dist/extensions/authoring.js +71 -0
- package/dist/extensions/authoring.js.map +1 -0
- package/dist/extensions/discovery.d.ts +3 -0
- package/dist/extensions/discovery.d.ts.map +1 -0
- package/dist/extensions/discovery.js +51 -0
- package/dist/extensions/discovery.js.map +1 -0
- package/dist/extensions/review.d.ts +3 -0
- package/dist/extensions/review.d.ts.map +1 -0
- package/dist/extensions/review.js +44 -0
- package/dist/extensions/review.js.map +1 -0
- package/dist/github.d.ts +3 -0
- package/dist/github.d.ts.map +1 -0
- package/dist/github.js +19 -0
- package/dist/github.js.map +1 -0
- package/dist/harbor-results.d.ts +14 -0
- package/dist/harbor-results.d.ts.map +1 -0
- package/dist/harbor-results.js +105 -0
- package/dist/harbor-results.js.map +1 -0
- package/dist/harbor-task.d.ts +12 -0
- package/dist/harbor-task.d.ts.map +1 -0
- package/dist/harbor-task.js +462 -0
- package/dist/harbor-task.js.map +1 -0
- package/dist/hash.d.ts +2 -0
- package/dist/hash.d.ts.map +1 -0
- package/dist/hash.js +5 -0
- package/dist/hash.js.map +1 -0
- package/dist/modal-auth.d.ts +2 -0
- package/dist/modal-auth.d.ts.map +1 -0
- package/dist/modal-auth.js +9 -0
- package/dist/modal-auth.js.map +1 -0
- package/dist/modal-executor.d.ts +12 -0
- package/dist/modal-executor.d.ts.map +1 -0
- package/dist/modal-executor.js +152 -0
- package/dist/modal-executor.js.map +1 -0
- package/dist/parallel.d.ts +2 -0
- package/dist/parallel.d.ts.map +1 -0
- package/dist/parallel.js +19 -0
- package/dist/parallel.js.map +1 -0
- package/dist/process.d.ts +27 -0
- package/dist/process.d.ts.map +1 -0
- package/dist/process.js +136 -0
- package/dist/process.js.map +1 -0
- package/dist/provenance.d.ts +22 -0
- package/dist/provenance.d.ts.map +1 -0
- package/dist/provenance.js +361 -0
- package/dist/provenance.js.map +1 -0
- package/dist/reaudit-main.d.ts +3 -0
- package/dist/reaudit-main.d.ts.map +1 -0
- package/dist/reaudit-main.js +155 -0
- package/dist/reaudit-main.js.map +1 -0
- package/dist/repair-main.d.ts +3 -0
- package/dist/repair-main.d.ts.map +1 -0
- package/dist/repair-main.js +168 -0
- package/dist/repair-main.js.map +1 -0
- package/dist/repair.d.ts +9 -0
- package/dist/repair.d.ts.map +1 -0
- package/dist/repair.js +46 -0
- package/dist/repair.js.map +1 -0
- package/dist/review/assets/abap-BdImnpbu.js +1 -0
- package/dist/review/assets/actionscript-3-B3316cI-.js +1 -0
- package/dist/review/assets/ada-bCR0ucgS.js +1 -0
- package/dist/review/assets/ahk-CsyLZFj1.js +1 -0
- package/dist/review/assets/ahk2-8Zs4aa1G.js +1 -0
- package/dist/review/assets/andromeeda-C4gqWexZ.js +1 -0
- package/dist/review/assets/angular-html-DA-rfuFy.js +1 -0
- package/dist/review/assets/angular-ts-BrjP3tb8.js +1 -0
- package/dist/review/assets/apache-Pmp26Uib.js +1 -0
- package/dist/review/assets/apex-DhZFqWV2.js +1 -0
- package/dist/review/assets/apl-CORt7UWP.js +1 -0
- package/dist/review/assets/applescript-Co6uUVPk.js +1 -0
- package/dist/review/assets/ara-BRHolxvo.js +1 -0
- package/dist/review/assets/asciidoc-CSVQ5wI8.js +1 -0
- package/dist/review/assets/asm-D_Q5rh1f.js +1 -0
- package/dist/review/assets/astro-HNnZUWAn.js +1 -0
- package/dist/review/assets/aurora-x-D-2ljcwZ.js +1 -0
- package/dist/review/assets/awk-DMzUqQB5.js +1 -0
- package/dist/review/assets/ayu-dark-DYE7WIF3.js +1 -0
- package/dist/review/assets/ayu-light-BA47KaF1.js +1 -0
- package/dist/review/assets/ayu-mirage-32ctXXKs.js +1 -0
- package/dist/review/assets/ballerina-BFfxhgS-.js +1 -0
- package/dist/review/assets/bat-CickPsom.js +1 -0
- package/dist/review/assets/beancount-k_qm7-4y.js +1 -0
- package/dist/review/assets/berry-uYugtg8r.js +1 -0
- package/dist/review/assets/bibtex-CHM0blh-.js +1 -0
- package/dist/review/assets/bicep-Bmn6On1c.js +1 -0
- package/dist/review/assets/bird2-Bx8U0n9b.js +1 -0
- package/dist/review/assets/blade-2xfisSek.js +1 -0
- package/dist/review/assets/bsl-DlhNcFeZ.js +1 -0
- package/dist/review/assets/c-BIGW1oBm.js +1 -0
- package/dist/review/assets/c3-Dp5svz6Z.js +1 -0
- package/dist/review/assets/cadence-Bv_4Rxtq.js +1 -0
- package/dist/review/assets/cairo-KRGpt6FW.js +1 -0
- package/dist/review/assets/catppuccin-frappe-CZL1YF0i.js +1 -0
- package/dist/review/assets/catppuccin-latte-DH-8KZSZ.js +1 -0
- package/dist/review/assets/catppuccin-macchiato-B7yYVSCf.js +1 -0
- package/dist/review/assets/catppuccin-mocha-Ct7hS0mc.js +1 -0
- package/dist/review/assets/chapel-DTp_pixX.js +1 -0
- package/dist/review/assets/clarity-Dn5IMItf.js +1 -0
- package/dist/review/assets/clojure-P80f7IUj.js +1 -0
- package/dist/review/assets/cmake-D1j8_8rp.js +1 -0
- package/dist/review/assets/cobol-nBiQ_Alo.js +1 -0
- package/dist/review/assets/codeowners-Bp6g37R7.js +1 -0
- package/dist/review/assets/codeql-DsOJ9woJ.js +1 -0
- package/dist/review/assets/coffee-Ch7k5sss.js +1 -0
- package/dist/review/assets/common-lisp-Cg-RD9OK.js +1 -0
- package/dist/review/assets/coq-C7JzOVbR.js +1 -0
- package/dist/review/assets/cpp-BMRokrvK.js +1 -0
- package/dist/review/assets/crystal-DGywbUpC.js +1 -0
- package/dist/review/assets/csharp-DSvCPggb.js +1 -0
- package/dist/review/assets/css-CLj8gQPS.js +1 -0
- package/dist/review/assets/csv-fuZLfV_i.js +1 -0
- package/dist/review/assets/cue-D82EKSYY.js +1 -0
- package/dist/review/assets/cypher-COkxafJQ.js +1 -0
- package/dist/review/assets/d-85-TOEBH.js +1 -0
- package/dist/review/assets/dark-plus-C3mMm8J8.js +1 -0
- package/dist/review/assets/dart-bE4Kk8sk.js +1 -0
- package/dist/review/assets/dax-CEL-wOlO.js +1 -0
- package/dist/review/assets/desktop-BmXAJ9_W.js +1 -0
- package/dist/review/assets/diff-D97Zzqfu.js +1 -0
- package/dist/review/assets/docker-BcOcwvcX.js +1 -0
- package/dist/review/assets/dotenv-Da5cRb03.js +1 -0
- package/dist/review/assets/dracula-BzJJZx-M.js +1 -0
- package/dist/review/assets/dracula-soft-BXkSAIEj.js +1 -0
- package/dist/review/assets/dream-maker-BtqSS_iP.js +1 -0
- package/dist/review/assets/edge-FbVlp4U3.js +1 -0
- package/dist/review/assets/elixir-CkH2-t6x.js +1 -0
- package/dist/review/assets/elm-DbKCFpqz.js +1 -0
- package/dist/review/assets/emacs-lisp-C_m_b--Z.js +1 -0
- package/dist/review/assets/erb-DXfck5VN.js +1 -0
- package/dist/review/assets/erlang-DsQrWhSR.js +1 -0
- package/dist/review/assets/everforest-dark-BgDCqdQA.js +1 -0
- package/dist/review/assets/everforest-light-C8M2exoo.js +1 -0
- package/dist/review/assets/fennel-BYunw83y.js +1 -0
- package/dist/review/assets/fish-BvzEVeQv.js +1 -0
- package/dist/review/assets/fluent-C4IJs8-o.js +1 -0
- package/dist/review/assets/fortran-fixed-form-CkoXwp7k.js +1 -0
- package/dist/review/assets/fortran-free-form-BxgE0vQu.js +1 -0
- package/dist/review/assets/fsharp-CXgrBDvD.js +1 -0
- package/dist/review/assets/gdresource-TyuKm33G.js +1 -0
- package/dist/review/assets/gdscript-DqcFQ5yU.js +1 -0
- package/dist/review/assets/gdshader-DkwncUOv.js +1 -0
- package/dist/review/assets/genie-D0YGMca9.js +1 -0
- package/dist/review/assets/gherkin-DyxjwDmM.js +1 -0
- package/dist/review/assets/git-commit-F4YmCXRG.js +1 -0
- package/dist/review/assets/git-rebase-r7XF79zn.js +1 -0
- package/dist/review/assets/github-dark-DHJKELXO.js +1 -0
- package/dist/review/assets/github-dark-default-Cuk6v7N8.js +1 -0
- package/dist/review/assets/github-dark-dimmed-DH5Ifo-i.js +1 -0
- package/dist/review/assets/github-dark-high-contrast-E3gJ1_iC.js +1 -0
- package/dist/review/assets/github-light-DAi9KRSo.js +1 -0
- package/dist/review/assets/github-light-default-D7oLnXFd.js +1 -0
- package/dist/review/assets/github-light-high-contrast-BfjtVDDH.js +1 -0
- package/dist/review/assets/gleam-BspZqrRM.js +1 -0
- package/dist/review/assets/glimmer-js-ByusRIyA.js +1 -0
- package/dist/review/assets/glimmer-ts-BfAWNZQY.js +1 -0
- package/dist/review/assets/glsl-DplSGwfg.js +1 -0
- package/dist/review/assets/gn-n2N0HUVH.js +1 -0
- package/dist/review/assets/gnuplot-DdkO51Og.js +1 -0
- package/dist/review/assets/go-C27-OAKa.js +1 -0
- package/dist/review/assets/graphql-ChdNCCLP.js +1 -0
- package/dist/review/assets/groovy-gcz8RCvz.js +1 -0
- package/dist/review/assets/gruvbox-dark-hard-CFHQjOhq.js +1 -0
- package/dist/review/assets/gruvbox-dark-medium-GsRaNv29.js +1 -0
- package/dist/review/assets/gruvbox-dark-soft-CVdnzihN.js +1 -0
- package/dist/review/assets/gruvbox-light-hard-CH1njM8p.js +1 -0
- package/dist/review/assets/gruvbox-light-medium-DRw_LuNl.js +1 -0
- package/dist/review/assets/gruvbox-light-soft-hJgmCMqR.js +1 -0
- package/dist/review/assets/hack-BWmVpMyf.js +1 -0
- package/dist/review/assets/haml-D5jkg6IW.js +1 -0
- package/dist/review/assets/handlebars-BpdQsYii.js +1 -0
- package/dist/review/assets/haskell-Df6bDoY_.js +1 -0
- package/dist/review/assets/haxe-CfZj7gIn.js +1 -0
- package/dist/review/assets/hcl-BWvSN4gD.js +1 -0
- package/dist/review/assets/hjson-D5-asLiD.js +1 -0
- package/dist/review/assets/hlsl-D3lLCCz7.js +1 -0
- package/dist/review/assets/horizon-BUw7H-hv.js +1 -0
- package/dist/review/assets/horizon-bright-CUuTKBJd.js +1 -0
- package/dist/review/assets/houston-DnULxvSX.js +1 -0
- package/dist/review/assets/html-derivative-DlHx6ybY.js +1 -0
- package/dist/review/assets/html-pp8916En.js +1 -0
- package/dist/review/assets/http-jrhK8wxY.js +1 -0
- package/dist/review/assets/hurl-irOxFIW8.js +1 -0
- package/dist/review/assets/hxml-2-FPmUDs.js +1 -0
- package/dist/review/assets/hy-DFXneXwc.js +1 -0
- package/dist/review/assets/imba-DGztddWO.js +1 -0
- package/dist/review/assets/index-Dq-6DFzI.css +1 -0
- package/dist/review/assets/index-DurW_RD7.js +1567 -0
- package/dist/review/assets/ini-BEwlwnbL.js +1 -0
- package/dist/review/assets/java-CylS5w8V.js +1 -0
- package/dist/review/assets/javascript-wDzz0qaB.js +1 -0
- package/dist/review/assets/jinja-f2NsQr07.js +1 -0
- package/dist/review/assets/jison-wvAkD_A8.js +1 -0
- package/dist/review/assets/json-Cp-IABpG.js +1 -0
- package/dist/review/assets/json5-C9tS-k6U.js +1 -0
- package/dist/review/assets/jsonc-Des-eS-w.js +1 -0
- package/dist/review/assets/jsonl-DcaNXYhu.js +1 -0
- package/dist/review/assets/jsonnet-DFQXde-d.js +1 -0
- package/dist/review/assets/jssm-C2t-YnRu.js +1 -0
- package/dist/review/assets/jsx-g9-lgVsj.js +1 -0
- package/dist/review/assets/julia-5Bft2YPA.js +1 -0
- package/dist/review/assets/just-Cwhn7H3k.js +1 -0
- package/dist/review/assets/kanagawa-dragon-BuwD2xS4.js +1 -0
- package/dist/review/assets/kanagawa-lotus-C3DzagqV.js +1 -0
- package/dist/review/assets/kanagawa-wave-DFGoQZhC.js +1 -0
- package/dist/review/assets/kdl-DV7GczEv.js +1 -0
- package/dist/review/assets/kotlin-BdnUsdx6.js +1 -0
- package/dist/review/assets/kusto-wEQ09or8.js +1 -0
- package/dist/review/assets/laserwave-DUszq2jm.js +1 -0
- package/dist/review/assets/latex-D5pSuvFb.js +1 -0
- package/dist/review/assets/lean-BZvkOJ9d.js +1 -0
- package/dist/review/assets/less-B1dDrJ26.js +1 -0
- package/dist/review/assets/light-plus-B7mTdjB0.js +1 -0
- package/dist/review/assets/liquid-C0sCDyMI.js +1 -0
- package/dist/review/assets/llvm-BZoOZj88.js +1 -0
- package/dist/review/assets/log-2UxHyX5q.js +1 -0
- package/dist/review/assets/logo-BtOb2qkB.js +1 -0
- package/dist/review/assets/lua-BaeVxFsk.js +1 -0
- package/dist/review/assets/luau-BnpPk5vE.js +1 -0
- package/dist/review/assets/make-CHLpvVh8.js +1 -0
- package/dist/review/assets/markdown-Cvjx9yec.js +1 -0
- package/dist/review/assets/marko-DjSrsDqO.js +1 -0
- package/dist/review/assets/material-theme-D5KoaKCx.js +1 -0
- package/dist/review/assets/material-theme-darker-BfHTSMKl.js +1 -0
- package/dist/review/assets/material-theme-lighter-B0m2ddpp.js +1 -0
- package/dist/review/assets/material-theme-ocean-CyktbL80.js +1 -0
- package/dist/review/assets/material-theme-palenight-Csfq5Kiy.js +1 -0
- package/dist/review/assets/matlab-D7o27uSR.js +1 -0
- package/dist/review/assets/mdc-D1_yUvq7.js +1 -0
- package/dist/review/assets/mdx-Cmh6b_Ma.js +1 -0
- package/dist/review/assets/mermaid-CQcHuHx7.js +1 -0
- package/dist/review/assets/min-dark-CafNBF8u.js +1 -0
- package/dist/review/assets/min-light-CTRr51gU.js +1 -0
- package/dist/review/assets/mipsasm-CKIfxQSi.js +1 -0
- package/dist/review/assets/mojo-Blby8dQK.js +1 -0
- package/dist/review/assets/monokai-BRVnQi9A.js +1 -0
- package/dist/review/assets/moonbit-CHtswR0a.js +1 -0
- package/dist/review/assets/move-el3G9tDJ.js +1 -0
- package/dist/review/assets/narrat-DRg8JJMk.js +1 -0
- package/dist/review/assets/nextflow-C-mBbutL.js +1 -0
- package/dist/review/assets/nextflow-groovy-vE_lwT2v.js +1 -0
- package/dist/review/assets/nginx-BpAMiNFr.js +1 -0
- package/dist/review/assets/night-owl-C39BiMTA.js +1 -0
- package/dist/review/assets/night-owl-light-CMTm3GFP.js +1 -0
- package/dist/review/assets/nim-BIad80T-.js +1 -0
- package/dist/review/assets/nix-CwoSXNpI.js +1 -0
- package/dist/review/assets/nord-Ddv68eIx.js +1 -0
- package/dist/review/assets/nsis-BlV79W_Q.js +1 -0
- package/dist/review/assets/nushell-D3jzshHO.js +1 -0
- package/dist/review/assets/objective-c-DXmwc3jG.js +1 -0
- package/dist/review/assets/objective-cpp-CLxacb5B.js +1 -0
- package/dist/review/assets/ocaml-C0hk2d4L.js +1 -0
- package/dist/review/assets/odin-BBf5iR-q.js +1 -0
- package/dist/review/assets/one-dark-pro-DVMEJ2y_.js +1 -0
- package/dist/review/assets/one-light-C3Wv6jpd.js +1 -0
- package/dist/review/assets/openscad-C4EeE6gA.js +1 -0
- package/dist/review/assets/org-DM6o9KBp.js +1 -0
- package/dist/review/assets/pascal-D93ZcfNL.js +1 -0
- package/dist/review/assets/perl-B9cMNwum.js +1 -0
- package/dist/review/assets/php-Csjmro_R.js +1 -0
- package/dist/review/assets/pierre-dark-CyvmCCZW.js +1 -0
- package/dist/review/assets/pierre-dark-protanopia-deuteranopia-Rgc0TwpF.js +1 -0
- package/dist/review/assets/pierre-dark-soft-BHGpRqa4.js +1 -0
- package/dist/review/assets/pierre-dark-tritanopia-Beq2gCRQ.js +1 -0
- package/dist/review/assets/pierre-dark-vibrant-BWBVywrn.js +1 -0
- package/dist/review/assets/pierre-light-480U9XYS.js +1 -0
- package/dist/review/assets/pierre-light-protanopia-deuteranopia-CaVOBURG.js +1 -0
- package/dist/review/assets/pierre-light-soft-CVdyfjmI.js +1 -0
- package/dist/review/assets/pierre-light-tritanopia-B4_gpKOM.js +1 -0
- package/dist/review/assets/pierre-light-vibrant-DdTDNdfJ.js +1 -0
- package/dist/review/assets/pkl-u5AG7uiY.js +1 -0
- package/dist/review/assets/plastic-3e1v2bzS.js +1 -0
- package/dist/review/assets/plsql-ChMvpjG-.js +1 -0
- package/dist/review/assets/po-BTJTHyun.js +1 -0
- package/dist/review/assets/poimandres-CS3Unz2-.js +1 -0
- package/dist/review/assets/polar-C0HS_06l.js +1 -0
- package/dist/review/assets/postcss-CXtECtnM.js +1 -0
- package/dist/review/assets/powerquery-CEu0bR-o.js +1 -0
- package/dist/review/assets/powershell-BmBUJMz7.js +1 -0
- package/dist/review/assets/prisma-Vru482bI.js +1 -0
- package/dist/review/assets/prolog-CbFg5uaA.js +1 -0
- package/dist/review/assets/proto-C7zT0LnQ.js +1 -0
- package/dist/review/assets/pug-DKIMFp6K.js +1 -0
- package/dist/review/assets/puppet-BMWR74SV.js +1 -0
- package/dist/review/assets/purescript-CklMAg4u.js +1 -0
- package/dist/review/assets/python-B6aJPvgy.js +1 -0
- package/dist/review/assets/qml-3beO22l8.js +1 -0
- package/dist/review/assets/qmldir-C8lEn-DE.js +1 -0
- package/dist/review/assets/qss-IeuSbFQv.js +1 -0
- package/dist/review/assets/r-Cf5RLm7j.js +1 -0
- package/dist/review/assets/racket-BqYA7rlc.js +1 -0
- package/dist/review/assets/raku-DXvB9xmW.js +1 -0
- package/dist/review/assets/razor-BjBPvh-w.js +1 -0
- package/dist/review/assets/rbs-CpoqiR4B.js +1 -0
- package/dist/review/assets/red-hvxz__6c.js +1 -0
- package/dist/review/assets/reg-C-SQnVFl.js +1 -0
- package/dist/review/assets/regexp-CDVJQ6XC.js +1 -0
- package/dist/review/assets/rel-C3B-1QV4.js +1 -0
- package/dist/review/assets/riscv-BM1_JUlF.js +1 -0
- package/dist/review/assets/ron-D8l8udqQ.js +1 -0
- package/dist/review/assets/rose-pine-5qJOZa0Y.js +1 -0
- package/dist/review/assets/rose-pine-dawn-zx0QlTCp.js +1 -0
- package/dist/review/assets/rose-pine-moon-C9pEdX9L.js +1 -0
- package/dist/review/assets/rosmsg-BJDFO7_C.js +1 -0
- package/dist/review/assets/rst-bs7f0vWN.js +1 -0
- package/dist/review/assets/ruby-C0TQ7zu5.js +1 -0
- package/dist/review/assets/rust-B1yitclQ.js +1 -0
- package/dist/review/assets/sas-DEy46yEz.js +1 -0
- package/dist/review/assets/sass-Cj5Yp3dK.js +1 -0
- package/dist/review/assets/scala-CqE71os6.js +1 -0
- package/dist/review/assets/scheme-C98Dy4si.js +1 -0
- package/dist/review/assets/scss-D5BDwBP9.js +1 -0
- package/dist/review/assets/sdbl-DVxCFoDh.js +1 -0
- package/dist/review/assets/shaderlab-Dg9Lc6iA.js +1 -0
- package/dist/review/assets/shellscript-Yzrsuije.js +1 -0
- package/dist/review/assets/shellsession-BADoaaVG.js +1 -0
- package/dist/review/assets/slack-dark-BthQWCQV.js +1 -0
- package/dist/review/assets/slack-ochin-DqwNpetd.js +1 -0
- package/dist/review/assets/smalltalk-BOQMe2GC.js +1 -0
- package/dist/review/assets/smithy-cds9vsN8.js +1 -0
- package/dist/review/assets/snazzy-light-Bw305WKR.js +1 -0
- package/dist/review/assets/solarized-dark-CpvCGNkr.js +1 -0
- package/dist/review/assets/solarized-light-Dlz6yCKv.js +1 -0
- package/dist/review/assets/solidity-DijEV5ha.js +1 -0
- package/dist/review/assets/soy-8wufbnw4.js +1 -0
- package/dist/review/assets/sparql-rVzFXLq3.js +1 -0
- package/dist/review/assets/splunk-BtCnVYZw.js +1 -0
- package/dist/review/assets/sql-CRqJ_cUM.js +1 -0
- package/dist/review/assets/ssh-config-_ykCGR6B.js +1 -0
- package/dist/review/assets/stata-DI20mbqo.js +1 -0
- package/dist/review/assets/stylus-BEDo0Tqx.js +1 -0
- package/dist/review/assets/surrealql-Cjom0U5J.js +1 -0
- package/dist/review/assets/svelte-Cy7k_4gC.js +1 -0
- package/dist/review/assets/swift-C2oV4EkX.js +1 -0
- package/dist/review/assets/synthwave-84-CbfX1IO0.js +1 -0
- package/dist/review/assets/system-verilog-0hqHdDBg.js +1 -0
- package/dist/review/assets/systemd-4A_iFExJ.js +1 -0
- package/dist/review/assets/talonscript-CkByrt1z.js +1 -0
- package/dist/review/assets/tasl-QIJgUcNo.js +1 -0
- package/dist/review/assets/tcl-dwOrl1Do.js +1 -0
- package/dist/review/assets/templ-DhtptRzy.js +1 -0
- package/dist/review/assets/terraform-BETggiCN.js +1 -0
- package/dist/review/assets/tex-D96PA37w.js +1 -0
- package/dist/review/assets/tokyo-night-hegEt444.js +1 -0
- package/dist/review/assets/toml-vGWfd6FD.js +1 -0
- package/dist/review/assets/ts-tags-D351s5mN.js +1 -0
- package/dist/review/assets/tsv-B_m7g4N7.js +1 -0
- package/dist/review/assets/tsx-COt5Ahok.js +1 -0
- package/dist/review/assets/turtle-BsS91CYL.js +1 -0
- package/dist/review/assets/twig-27uCiNez.js +1 -0
- package/dist/review/assets/typescript-BPQ3VLAy.js +1 -0
- package/dist/review/assets/typespec-CAFt9gP4.js +1 -0
- package/dist/review/assets/typst-DHCkPAjA.js +1 -0
- package/dist/review/assets/v-BGw2Nkan.js +1 -0
- package/dist/review/assets/vala-CsfeWuGM.js +1 -0
- package/dist/review/assets/vb-Cu-pLBUe.js +1 -0
- package/dist/review/assets/verilog-nZwndyjY.js +1 -0
- package/dist/review/assets/vesper-DRje8inN.js +1 -0
- package/dist/review/assets/vhdl-CeAyd5Ju.js +1 -0
- package/dist/review/assets/viml-CJc9bBzg.js +1 -0
- package/dist/review/assets/vitesse-black-Bkuqu6BP.js +1 -0
- package/dist/review/assets/vitesse-dark-D0r3Knsf.js +1 -0
- package/dist/review/assets/vitesse-light-CVO1_9PV.js +1 -0
- package/dist/review/assets/vue-BqiEGhQt.js +1 -0
- package/dist/review/assets/vue-html-AaS7Mt5G.js +1 -0
- package/dist/review/assets/vue-vine-BoDAl6tE.js +1 -0
- package/dist/review/assets/vyper-CDx5xZoG.js +1 -0
- package/dist/review/assets/wasm-CG6Dc4jp.js +1 -0
- package/dist/review/assets/wasm-MzD3tlZU.js +1 -0
- package/dist/review/assets/wenyan-BV7otONQ.js +1 -0
- package/dist/review/assets/wgsl-Dx-B1_4e.js +1 -0
- package/dist/review/assets/wikitext-BhOHFoWU.js +1 -0
- package/dist/review/assets/wit-5i3qLPDT.js +1 -0
- package/dist/review/assets/wolfram-lXgVvXCa.js +1 -0
- package/dist/review/assets/xml-sdJ4AIDG.js +1 -0
- package/dist/review/assets/xsl-CtQFsRM5.js +1 -0
- package/dist/review/assets/yaml-Buea-lGh.js +1 -0
- package/dist/review/assets/zenscript-DVFEvuxE.js +1 -0
- package/dist/review/assets/zig-VOosw3JB.js +1 -0
- package/dist/review/index.html +15 -0
- package/dist/run-wait.d.ts +13 -0
- package/dist/run-wait.d.ts.map +1 -0
- package/dist/run-wait.js +25 -0
- package/dist/run-wait.js.map +1 -0
- package/dist/sandbox-author.bundle.js +14979 -0
- package/dist/sandbox-author.d.ts +3 -0
- package/dist/sandbox-author.d.ts.map +1 -0
- package/dist/sandbox-author.js +17 -0
- package/dist/sandbox-author.js.map +1 -0
- package/dist/sandbox-repair.bundle.js +314 -0
- package/dist/sandbox-repair.d.ts +3 -0
- package/dist/sandbox-repair.d.ts.map +1 -0
- package/dist/sandbox-repair.js +121 -0
- package/dist/sandbox-repair.js.map +1 -0
- package/dist/sandbox-review.bundle.js +14458 -0
- package/dist/sandbox-review.d.ts +2 -0
- package/dist/sandbox-review.d.ts.map +1 -0
- package/dist/sandbox-review.js +16 -0
- package/dist/sandbox-review.js.map +1 -0
- package/dist/sandbox-validation-repair.bundle.js +14744 -0
- package/dist/sandbox-validation-repair.d.ts +3 -0
- package/dist/sandbox-validation-repair.d.ts.map +1 -0
- package/dist/sandbox-validation-repair.js +134 -0
- package/dist/sandbox-validation-repair.js.map +1 -0
- package/dist/sandbox.d.ts +39 -0
- package/dist/sandbox.d.ts.map +1 -0
- package/dist/sandbox.js +8 -0
- package/dist/sandbox.js.map +1 -0
- package/dist/subscription-auth.d.ts +6 -0
- package/dist/subscription-auth.d.ts.map +1 -0
- package/dist/subscription-auth.js +59 -0
- package/dist/subscription-auth.js.map +1 -0
- package/dist/temporal.d.ts +6 -0
- package/dist/temporal.d.ts.map +1 -0
- package/dist/temporal.js +17 -0
- package/dist/temporal.js.map +1 -0
- package/dist/validate-main.d.ts +3 -0
- package/dist/validate-main.d.ts.map +1 -0
- package/dist/validate-main.js +137 -0
- package/dist/validate-main.js.map +1 -0
- package/dist/validation-repair.d.ts +10 -0
- package/dist/validation-repair.d.ts.map +1 -0
- package/dist/validation-repair.js +72 -0
- package/dist/validation-repair.js.map +1 -0
- package/dist/worker-main.d.ts +2 -0
- package/dist/worker-main.d.ts.map +1 -0
- package/dist/worker-main.js +30 -0
- package/dist/worker-main.js.map +1 -0
- package/dist/workflow.d.ts +6 -0
- package/dist/workflow.d.ts.map +1 -0
- package/dist/workflow.js +428 -0
- package/dist/workflow.js.map +1 -0
- package/package.json +74 -0
- package/src/extensions/authoring.ts +86 -0
- package/src/extensions/discovery.ts +61 -0
- package/src/extensions/review.ts +54 -0
- package/src/skills/selfbench/SKILL.md +119 -0
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"sandbox-validation-repair.d.ts","sourceRoot":"","sources":["../src/sandbox-validation-repair.ts"],"names":[],"mappings":""}
|
|
@@ -0,0 +1,134 @@
|
|
|
1
|
+
#!/usr/bin/env node
|
|
2
|
+
import { access, chmod, mkdir, readFile, writeFile } from "node:fs/promises";
|
|
3
|
+
import { homedir, tmpdir } from "node:os";
|
|
4
|
+
import { join } from "node:path";
|
|
5
|
+
import { taskDefinitionSchema } from "./contracts.js";
|
|
6
|
+
import { runCommand } from "./process.js";
|
|
7
|
+
import { assertValidationRepair, validationRepairPaths, validationRepairPrompt, } from "./validation-repair.js";
|
|
8
|
+
const [archivePath, definitionPath, diagnosticsPath, outputDefinition, outputPatch, outputReport] = process.argv.slice(2);
|
|
9
|
+
if (!archivePath ||
|
|
10
|
+
!definitionPath ||
|
|
11
|
+
!diagnosticsPath ||
|
|
12
|
+
!outputDefinition ||
|
|
13
|
+
!outputPatch ||
|
|
14
|
+
!outputReport) {
|
|
15
|
+
throw new Error("usage: sandbox-validation-repair TASK.tar.gz DEFINITION.json DIAGNOSTICS.txt OUTPUT-DEFINITION.json OUTPUT-PATCH OUTPUT-REPORT.json");
|
|
16
|
+
}
|
|
17
|
+
const extractedDirectory = "/work/task";
|
|
18
|
+
const repositoryDirectory = "/work/repo";
|
|
19
|
+
await Promise.all([mkdir(extractedDirectory, { recursive: true }), mkdir(repositoryDirectory)]);
|
|
20
|
+
await runCommand("tar", ["-xzf", archivePath, "-C", extractedDirectory]);
|
|
21
|
+
const taskDirectory = await access(join(extractedDirectory, "instruction.md")).then(() => extractedDirectory, () => join(extractedDirectory, "harbor-task"));
|
|
22
|
+
const [instruction, originalPatch, diagnostics, originalDefinitionBytes] = await Promise.all([
|
|
23
|
+
readFile(join(taskDirectory, "instruction.md"), "utf8"),
|
|
24
|
+
readFile(join(taskDirectory, "tests/test.patch"), "utf8"),
|
|
25
|
+
readFile(diagnosticsPath, "utf8"),
|
|
26
|
+
readFile(definitionPath, "utf8"),
|
|
27
|
+
]);
|
|
28
|
+
const originalDefinition = taskDefinitionSchema.parse(JSON.parse(originalDefinitionBytes));
|
|
29
|
+
const allowedPaths = validationRepairPaths(originalPatch);
|
|
30
|
+
await runCommand("tar", [
|
|
31
|
+
"-xzf",
|
|
32
|
+
join(taskDirectory, "tests/repo.tar.gz"),
|
|
33
|
+
"-C",
|
|
34
|
+
repositoryDirectory,
|
|
35
|
+
]);
|
|
36
|
+
await runCommand("git", ["-C", repositoryDirectory, "init", "-q"]);
|
|
37
|
+
await runCommand("git", ["-C", repositoryDirectory, "config", "user.name", "SelfBench"]);
|
|
38
|
+
await runCommand("git", ["-C", repositoryDirectory, "config", "user.email", "selfbench@local"]);
|
|
39
|
+
await runCommand("git", ["-C", repositoryDirectory, "add", "-A"]);
|
|
40
|
+
await runCommand("git", ["-C", repositoryDirectory, "commit", "-qm", "base"]);
|
|
41
|
+
await runCommand("git", [
|
|
42
|
+
"-C",
|
|
43
|
+
repositoryDirectory,
|
|
44
|
+
"apply",
|
|
45
|
+
join(taskDirectory, "tests/test.patch"),
|
|
46
|
+
]);
|
|
47
|
+
await runCommand("git", ["-C", repositoryDirectory, "add", "-N", "--all"]);
|
|
48
|
+
await Promise.all([
|
|
49
|
+
writeFile("/work/definition.json", `${JSON.stringify(originalDefinition, null, 2)}\n`),
|
|
50
|
+
writeFile("/work/gold.patch", await readFile(join(taskDirectory, "solution/gold.patch"))),
|
|
51
|
+
writeFile("/work/test.sh", await readFile(join(taskDirectory, "tests/test.sh"))),
|
|
52
|
+
]);
|
|
53
|
+
const piHome = join(homedir(), ".pi/agent");
|
|
54
|
+
const piAuth = process.env.SELFBENCH_PI_AUTH_JSON ?? fail("SELFBENCH_PI_AUTH_JSON is required");
|
|
55
|
+
await mkdir(piHome, { recursive: true });
|
|
56
|
+
await Promise.all([
|
|
57
|
+
writeFile(join(piHome, "auth.json"), piAuth).then(() => chmod(join(piHome, "auth.json"), 0o600)),
|
|
58
|
+
writeFile(join(piHome, "settings.json"), `${JSON.stringify({ transport: "auto" })}\n`),
|
|
59
|
+
]);
|
|
60
|
+
const promptPath = join(tmpdir(), `selfbench-validation-repair-${originalDefinition.taskId}.md`);
|
|
61
|
+
await writeFile(promptPath, validationRepairPrompt({
|
|
62
|
+
definition: originalDefinition,
|
|
63
|
+
authenticRequest: instruction,
|
|
64
|
+
diagnostics,
|
|
65
|
+
allowedPaths,
|
|
66
|
+
}));
|
|
67
|
+
const pi = await runCommand("pi", [
|
|
68
|
+
"--print",
|
|
69
|
+
"--mode",
|
|
70
|
+
"json",
|
|
71
|
+
"--no-session",
|
|
72
|
+
"--no-approve",
|
|
73
|
+
"--no-skills",
|
|
74
|
+
"--no-prompt-templates",
|
|
75
|
+
"--no-context-files",
|
|
76
|
+
"--no-extensions",
|
|
77
|
+
"--provider",
|
|
78
|
+
"openai-codex",
|
|
79
|
+
"--model",
|
|
80
|
+
process.env.SELFBENCH_REPAIR_MODEL ?? "gpt-5.6-sol",
|
|
81
|
+
"--thinking",
|
|
82
|
+
"high",
|
|
83
|
+
"--tools",
|
|
84
|
+
"read,bash,grep,find,ls",
|
|
85
|
+
`@${promptPath}`,
|
|
86
|
+
], {
|
|
87
|
+
allowFailure: true,
|
|
88
|
+
timeoutMs: 90 * 60 * 1000,
|
|
89
|
+
cwd: repositoryDirectory,
|
|
90
|
+
env: withoutApiKey(process.env),
|
|
91
|
+
onOutput: (stream, chunk) => {
|
|
92
|
+
(stream === "stdout" ? process.stdout : process.stderr).write(chunk);
|
|
93
|
+
},
|
|
94
|
+
});
|
|
95
|
+
if (pi.exitCode !== 0) {
|
|
96
|
+
throw new Error(`Pi validation repair exited ${pi.exitCode}: ${pi.stderr.slice(-2_000)}`);
|
|
97
|
+
}
|
|
98
|
+
const repairedDefinition = taskDefinitionSchema.parse(JSON.parse(await readFile("/work/definition.json", "utf8")));
|
|
99
|
+
const [tracked, untracked] = await Promise.all([
|
|
100
|
+
runCommand("git", ["-C", repositoryDirectory, "diff", "--name-only", "HEAD"]),
|
|
101
|
+
runCommand("git", ["-C", repositoryDirectory, "ls-files", "--others", "--exclude-standard"]),
|
|
102
|
+
]);
|
|
103
|
+
const changedPaths = [...tracked.stdout.split("\n"), ...untracked.stdout.split("\n")]
|
|
104
|
+
.filter(Boolean)
|
|
105
|
+
.sort();
|
|
106
|
+
assertValidationRepair(originalDefinition, repairedDefinition, originalPatch, changedPaths);
|
|
107
|
+
const repaired = await runCommand("git", ["-C", repositoryDirectory, "diff", "--binary", "HEAD"]);
|
|
108
|
+
const repairedPatch = repaired.stdout.startsWith("diff --git ") ? repaired.stdout : originalPatch;
|
|
109
|
+
if (repairedPatch === originalPatch &&
|
|
110
|
+
JSON.stringify(repairedDefinition) === JSON.stringify(originalDefinition)) {
|
|
111
|
+
throw new Error("validation repair left the task unchanged");
|
|
112
|
+
}
|
|
113
|
+
await Promise.all([
|
|
114
|
+
writeFile(outputDefinition, `${JSON.stringify(repairedDefinition, null, 2)}\n`),
|
|
115
|
+
writeFile(outputPatch, repairedPatch),
|
|
116
|
+
writeFile(outputReport, `${JSON.stringify({
|
|
117
|
+
schemaVersion: 1,
|
|
118
|
+
taskId: originalDefinition.taskId,
|
|
119
|
+
model: process.env.SELFBENCH_REPAIR_MODEL ?? "gpt-5.6-sol",
|
|
120
|
+
changedPaths,
|
|
121
|
+
definitionChanged: JSON.stringify(repairedDefinition) !== JSON.stringify(originalDefinition),
|
|
122
|
+
testsChanged: repairedPatch !== originalPatch,
|
|
123
|
+
piOutputTail: pi.stdout.slice(-4_000),
|
|
124
|
+
}, null, 2)}\n`),
|
|
125
|
+
]);
|
|
126
|
+
function withoutApiKey(environment) {
|
|
127
|
+
const output = { ...environment };
|
|
128
|
+
delete output.OPENAI_API_KEY;
|
|
129
|
+
return output;
|
|
130
|
+
}
|
|
131
|
+
function fail(message) {
|
|
132
|
+
throw new Error(message);
|
|
133
|
+
}
|
|
134
|
+
//# sourceMappingURL=sandbox-validation-repair.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"sandbox-validation-repair.js","sourceRoot":"","sources":["../src/sandbox-validation-repair.ts"],"names":[],"mappings":";AAEA,OAAO,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,QAAQ,EAAE,SAAS,EAAE,MAAM,kBAAkB,CAAC;AAC7E,OAAO,EAAE,OAAO,EAAE,MAAM,EAAE,MAAM,SAAS,CAAC;AAC1C,OAAO,EAAE,IAAI,EAAE,MAAM,WAAW,CAAC;AACjC,OAAO,EAAE,oBAAoB,EAAE,MAAM,gBAAgB,CAAC;AACtD,OAAO,EAAE,UAAU,EAAE,MAAM,cAAc,CAAC;AAC1C,OAAO,EACL,sBAAsB,EACtB,qBAAqB,EACrB,sBAAsB,GACvB,MAAM,wBAAwB,CAAC;AAEhC,MAAM,CAAC,WAAW,EAAE,cAAc,EAAE,eAAe,EAAE,gBAAgB,EAAE,WAAW,EAAE,YAAY,CAAC,GAC/F,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;AACxB,IACE,CAAC,WAAW;IACZ,CAAC,cAAc;IACf,CAAC,eAAe;IAChB,CAAC,gBAAgB;IACjB,CAAC,WAAW;IACZ,CAAC,YAAY,EACb,CAAC;IACD,MAAM,IAAI,KAAK,CACb,qIAAqI,CACtI,CAAC;AACJ,CAAC;AAED,MAAM,kBAAkB,GAAG,YAAY,CAAC;AACxC,MAAM,mBAAmB,GAAG,YAAY,CAAC;AACzC,MAAM,OAAO,CAAC,GAAG,CAAC,CAAC,KAAK,CAAC,kBAAkB,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,EAAE,KAAK,CAAC,mBAAmB,CAAC,CAAC,CAAC,CAAC;AAChG,MAAM,UAAU,CAAC,KAAK,EAAE,CAAC,MAAM,EAAE,WAAW,EAAE,IAAI,EAAE,kBAAkB,CAAC,CAAC,CAAC;AACzE,MAAM,aAAa,GAAG,MAAM,MAAM,CAAC,IAAI,CAAC,kBAAkB,EAAE,gBAAgB,CAAC,CAAC,CAAC,IAAI,CACjF,GAAG,EAAE,CAAC,kBAAkB,EACxB,GAAG,EAAE,CAAC,IAAI,CAAC,kBAAkB,EAAE,aAAa,CAAC,CAC9C,CAAC;AACF,MAAM,CAAC,WAAW,EAAE,aAAa,EAAE,WAAW,EAAE,uBAAuB,CAAC,GAAG,MAAM,OAAO,CAAC,GAAG,CAAC;IAC3F,QAAQ,CAAC,IAAI,CAAC,aAAa,EAAE,gBAAgB,CAAC,EAAE,MAAM,CAAC;IACvD,QAAQ,CAAC,IAAI,CAAC,aAAa,EAAE,kBAAkB,CAAC,EAAE,MAAM,CAAC;IACzD,QAAQ,CAAC,eAAe,EAAE,MAAM,CAAC;IACjC,QAAQ,CAAC,cAAc,EAAE,MAAM,CAAC;CACjC,CAAC,CAAC;AACH,MAAM,kBAAkB,GAAG,oBAAoB,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,uBAAuB,CAAC,CAAC,CAAC;AAC3F,MAAM,YAAY,GAAG,qBAAqB,CAAC,aAAa,CAAC,CAAC;AAE1D,MAAM,UAAU,CAAC,KAAK,EAAE;IACtB,MAAM;IACN,IAAI,CAAC,aAAa,EAAE,mBAAmB,CAAC;IACxC,IAAI;IACJ,mBAAmB;CACpB,CAAC,CAAC;AACH,MAAM,UAAU,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,mBAAmB,EAAE,MAAM,EAAE,IAAI,CAAC,CAAC,CAAC;AACnE,MAAM,UAAU,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,mBAAmB,EAAE,QAAQ,EAAE,WAAW,EAAE,WAAW,CAAC,CAAC,CAAC;AACzF,MAAM,UAAU,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,mBAAmB,EAAE,QAAQ,EAAE,YAAY,EAAE,iBAAiB,CAAC,CAAC,CAAC;AAChG,MAAM,UAAU,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,mBAAmB,EAAE,KAAK,EAAE,IAAI,CAAC,CAAC,CAAC;AAClE,MAAM,UAAU,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,mBAAmB,EAAE,QAAQ,EAAE,KAAK,EAAE,MAAM,CAAC,CAAC,CAAC;AAC9E,MAAM,UAAU,CAAC,KAAK,EAAE;IACtB,IAAI;IACJ,mBAAmB;IACnB,OAAO;IACP,IAAI,CAAC,aAAa,EAAE,kBAAkB,CAAC;CACxC,CAAC,CAAC;AACH,MAAM,UAAU,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,mBAAmB,EAAE,KAAK,EAAE,IAAI,EAAE,OAAO,CAAC,CAAC,CAAC;AAC3E,MAAM,OAAO,CAAC,GAAG,CAAC;IAChB,SAAS,CAAC,uBAAuB,EAAE,GAAG,IAAI,CAAC,SAAS,CAAC,kBAAkB,EAAE,IAAI,EAAE,CAAC,CAAC,IAAI,CAAC;IACtF,SAAS,CAAC,kBAAkB,EAAE,MAAM,QAAQ,CAAC,IAAI,CAAC,aAAa,EAAE,qBAAqB,CAAC,CAAC,CAAC;IACzF,SAAS,CAAC,eAAe,EAAE,MAAM,QAAQ,CAAC,IAAI,CAAC,aAAa,EAAE,eAAe,CAAC,CAAC,CAAC;CACjF,CAAC,CAAC;AAEH,MAAM,MAAM,GAAG,IAAI,CAAC,OAAO,EAAE,EAAE,WAAW,CAAC,CAAC;AAC5C,MAAM,MAAM,GAAG,OAAO,CAAC,GAAG,CAAC,sBAAsB,IAAI,IAAI,CAAC,oCAAoC,CAAC,CAAC;AAChG,MAAM,KAAK,CAAC,MAAM,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;AACzC,MAAM,OAAO,CAAC,GAAG,CAAC;IAChB,SAAS,CAAC,IAAI,CAAC,MAAM,EAAE,WAAW,CAAC,EAAE,MAAM,CAAC,CAAC,IAAI,CAAC,GAAG,EAAE,CAAC,KAAK,CAAC,IAAI,CAAC,MAAM,EAAE,WAAW,CAAC,EAAE,KAAK,CAAC,CAAC;IAChG,SAAS,CAAC,IAAI,CAAC,MAAM,EAAE,eAAe,CAAC,EAAE,GAAG,IAAI,CAAC,SAAS,CAAC,EAAE,SAAS,EAAE,MAAM,EAAE,CAAC,IAAI,CAAC;CACvF,CAAC,CAAC;AACH,MAAM,UAAU,GAAG,IAAI,CAAC,MAAM,EAAE,EAAE,+BAA+B,kBAAkB,CAAC,MAAM,KAAK,CAAC,CAAC;AACjG,MAAM,SAAS,CACb,UAAU,EACV,sBAAsB,CAAC;IACrB,UAAU,EAAE,kBAAkB;IAC9B,gBAAgB,EAAE,WAAW;IAC7B,WAAW;IACX,YAAY;CACb,CAAC,CACH,CAAC;AAEF,MAAM,EAAE,GAAG,MAAM,UAAU,CACzB,IAAI,EACJ;IACE,SAAS;IACT,QAAQ;IACR,MAAM;IACN,cAAc;IACd,cAAc;IACd,aAAa;IACb,uBAAuB;IACvB,oBAAoB;IACpB,iBAAiB;IACjB,YAAY;IACZ,cAAc;IACd,SAAS;IACT,OAAO,CAAC,GAAG,CAAC,sBAAsB,IAAI,aAAa;IACnD,YAAY;IACZ,MAAM;IACN,SAAS;IACT,wBAAwB;IACxB,IAAI,UAAU,EAAE;CACjB,EACD;IACE,YAAY,EAAE,IAAI;IAClB,SAAS,EAAE,EAAE,GAAG,EAAE,GAAG,IAAI;IACzB,GAAG,EAAE,mBAAmB;IACxB,GAAG,EAAE,aAAa,CAAC,OAAO,CAAC,GAAG,CAAC;IAC/B,QAAQ,EAAE,CAAC,MAAM,EAAE,KAAK,EAAE,EAAE;QAC1B,CAAC,MAAM,KAAK,QAAQ,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;IACvE,CAAC;CACF,CACF,CAAC;AACF,IAAI,EAAE,CAAC,QAAQ,KAAK,CAAC,EAAE,CAAC;IACtB,MAAM,IAAI,KAAK,CAAC,+BAA+B,EAAE,CAAC,QAAQ,KAAK,EAAE,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC;AAC5F,CAAC;AAED,MAAM,kBAAkB,GAAG,oBAAoB,CAAC,KAAK,CACnD,IAAI,CAAC,KAAK,CAAC,MAAM,QAAQ,CAAC,uBAAuB,EAAE,MAAM,CAAC,CAAC,CAC5D,CAAC;AACF,MAAM,CAAC,OAAO,EAAE,SAAS,CAAC,GAAG,MAAM,OAAO,CAAC,GAAG,CAAC;IAC7C,UAAU,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,mBAAmB,EAAE,MAAM,EAAE,aAAa,EAAE,MAAM,CAAC,CAAC;IAC7E,UAAU,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,mBAAmB,EAAE,UAAU,EAAE,UAAU,EAAE,oBAAoB,CAAC,CAAC;CAC7F,CAAC,CAAC;AACH,MAAM,YAAY,GAAG,CAAC,GAAG,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,EAAE,GAAG,SAAS,CAAC,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;KAClF,MAAM,CAAC,OAAO,CAAC;KACf,IAAI,EAAE,CAAC;AACV,sBAAsB,CAAC,kBAAkB,EAAE,kBAAkB,EAAE,aAAa,EAAE,YAAY,CAAC,CAAC;AAC5F,MAAM,QAAQ,GAAG,MAAM,UAAU,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,mBAAmB,EAAE,MAAM,EAAE,UAAU,EAAE,MAAM,CAAC,CAAC,CAAC;AAClG,MAAM,aAAa,GAAG,QAAQ,CAAC,MAAM,CAAC,UAAU,CAAC,aAAa,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC,CAAC,aAAa,CAAC;AAClG,IACE,aAAa,KAAK,aAAa;IAC/B,IAAI,CAAC,SAAS,CAAC,kBAAkB,CAAC,KAAK,IAAI,CAAC,SAAS,CAAC,kBAAkB,CAAC,EACzE,CAAC;IACD,MAAM,IAAI,KAAK,CAAC,2CAA2C,CAAC,CAAC;AAC/D,CAAC;AAED,MAAM,OAAO,CAAC,GAAG,CAAC;IAChB,SAAS,CAAC,gBAAgB,EAAE,GAAG,IAAI,CAAC,SAAS,CAAC,kBAAkB,EAAE,IAAI,EAAE,CAAC,CAAC,IAAI,CAAC;IAC/E,SAAS,CAAC,WAAW,EAAE,aAAa,CAAC;IACrC,SAAS,CACP,YAAY,EACZ,GAAG,IAAI,CAAC,SAAS,CACf;QACE,aAAa,EAAE,CAAC;QAChB,MAAM,EAAE,kBAAkB,CAAC,MAAM;QACjC,KAAK,EAAE,OAAO,CAAC,GAAG,CAAC,sBAAsB,IAAI,aAAa;QAC1D,YAAY;QACZ,iBAAiB,EACf,IAAI,CAAC,SAAS,CAAC,kBAAkB,CAAC,KAAK,IAAI,CAAC,SAAS,CAAC,kBAAkB,CAAC;QAC3E,YAAY,EAAE,aAAa,KAAK,aAAa;QAC7C,YAAY,EAAE,EAAE,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,KAAK,CAAC;KACtC,EACD,IAAI,EACJ,CAAC,CACF,IAAI,CACN;CACF,CAAC,CAAC;AAEH,SAAS,aAAa,CAAC,WAA8B;IACnD,MAAM,MAAM,GAAG,EAAE,GAAG,WAAW,EAAE,CAAC;IAClC,OAAO,MAAM,CAAC,cAAc,CAAC;IAC7B,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,SAAS,IAAI,CAAC,OAAe;IAC3B,MAAM,IAAI,KAAK,CAAC,OAAO,CAAC,CAAC;AAC3B,CAAC"}
|
|
@@ -0,0 +1,39 @@
|
|
|
1
|
+
import type { SelfBenchConfig } from "./config.js";
|
|
2
|
+
export interface SandboxFile {
|
|
3
|
+
readonly path: string;
|
|
4
|
+
readonly contents: Uint8Array | string;
|
|
5
|
+
}
|
|
6
|
+
export interface SandboxRequest {
|
|
7
|
+
readonly runId: string;
|
|
8
|
+
readonly stage: string;
|
|
9
|
+
readonly command: readonly string[];
|
|
10
|
+
readonly files?: readonly SandboxFile[];
|
|
11
|
+
readonly outputPaths?: readonly string[];
|
|
12
|
+
readonly environment?: Readonly<Record<string, string>>;
|
|
13
|
+
readonly secrets?: Readonly<Record<string, string>>;
|
|
14
|
+
readonly timeoutMs: number;
|
|
15
|
+
readonly inactivityTimeoutMs?: number;
|
|
16
|
+
readonly cpu?: number;
|
|
17
|
+
readonly memoryMiB?: number;
|
|
18
|
+
}
|
|
19
|
+
export interface SandboxProgress {
|
|
20
|
+
readonly stream: "stdout" | "stderr";
|
|
21
|
+
readonly bytes: number;
|
|
22
|
+
}
|
|
23
|
+
export interface SandboxRunOptions {
|
|
24
|
+
readonly signal?: AbortSignal;
|
|
25
|
+
readonly onProgress?: (progress: SandboxProgress) => void;
|
|
26
|
+
}
|
|
27
|
+
export interface SandboxResult {
|
|
28
|
+
readonly sandboxId: string;
|
|
29
|
+
readonly exitCode: number;
|
|
30
|
+
readonly stdout: string;
|
|
31
|
+
readonly stderr: string;
|
|
32
|
+
readonly outputs: Readonly<Record<string, Uint8Array>>;
|
|
33
|
+
}
|
|
34
|
+
export interface SandboxExecutor {
|
|
35
|
+
run(request: SandboxRequest, options?: SandboxRunOptions): Promise<SandboxResult>;
|
|
36
|
+
close(): void;
|
|
37
|
+
}
|
|
38
|
+
export declare function createSandboxExecutor(config: SelfBenchConfig["execution"]): SandboxExecutor;
|
|
39
|
+
//# sourceMappingURL=sandbox.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"sandbox.d.ts","sourceRoot":"","sources":["../src/sandbox.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,aAAa,CAAC;AAInD,MAAM,WAAW,WAAW;IAC1B,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC;IACtB,QAAQ,CAAC,QAAQ,EAAE,UAAU,GAAG,MAAM,CAAC;CACxC;AAED,MAAM,WAAW,cAAc;IAC7B,QAAQ,CAAC,KAAK,EAAE,MAAM,CAAC;IACvB,QAAQ,CAAC,KAAK,EAAE,MAAM,CAAC;IACvB,QAAQ,CAAC,OAAO,EAAE,SAAS,MAAM,EAAE,CAAC;IACpC,QAAQ,CAAC,KAAK,CAAC,EAAE,SAAS,WAAW,EAAE,CAAC;IACxC,QAAQ,CAAC,WAAW,CAAC,EAAE,SAAS,MAAM,EAAE,CAAC;IACzC,QAAQ,CAAC,WAAW,CAAC,EAAE,QAAQ,CAAC,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC,CAAC;IACxD,QAAQ,CAAC,OAAO,CAAC,EAAE,QAAQ,CAAC,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC,CAAC;IACpD,QAAQ,CAAC,SAAS,EAAE,MAAM,CAAC;IAC3B,QAAQ,CAAC,mBAAmB,CAAC,EAAE,MAAM,CAAC;IACtC,QAAQ,CAAC,GAAG,CAAC,EAAE,MAAM,CAAC;IACtB,QAAQ,CAAC,SAAS,CAAC,EAAE,MAAM,CAAC;CAC7B;AAED,MAAM,WAAW,eAAe;IAC9B,QAAQ,CAAC,MAAM,EAAE,QAAQ,GAAG,QAAQ,CAAC;IACrC,QAAQ,CAAC,KAAK,EAAE,MAAM,CAAC;CACxB;AAED,MAAM,WAAW,iBAAiB;IAChC,QAAQ,CAAC,MAAM,CAAC,EAAE,WAAW,CAAC;IAC9B,QAAQ,CAAC,UAAU,CAAC,EAAE,CAAC,QAAQ,EAAE,eAAe,KAAK,IAAI,CAAC;CAC3D;AAED,MAAM,WAAW,aAAa;IAC5B,QAAQ,CAAC,SAAS,EAAE,MAAM,CAAC;IAC3B,QAAQ,CAAC,QAAQ,EAAE,MAAM,CAAC;IAC1B,QAAQ,CAAC,MAAM,EAAE,MAAM,CAAC;IACxB,QAAQ,CAAC,MAAM,EAAE,MAAM,CAAC;IACxB,QAAQ,CAAC,OAAO,EAAE,QAAQ,CAAC,MAAM,CAAC,MAAM,EAAE,UAAU,CAAC,CAAC,CAAC;CACxD;AAED,MAAM,WAAW,eAAe;IAC9B,GAAG,CAAC,OAAO,EAAE,cAAc,EAAE,OAAO,CAAC,EAAE,iBAAiB,GAAG,OAAO,CAAC,aAAa,CAAC,CAAC;IAClF,KAAK,IAAI,IAAI,CAAC;CACf;AAED,wBAAgB,qBAAqB,CAAC,MAAM,EAAE,eAAe,CAAC,WAAW,CAAC,GAAG,eAAe,CAI3F"}
|
package/dist/sandbox.js
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
1
|
+
import { DockerSandboxExecutor } from "./docker-executor.js";
|
|
2
|
+
import { ModalSandboxExecutor } from "./modal-executor.js";
|
|
3
|
+
export function createSandboxExecutor(config) {
|
|
4
|
+
return config.kind === "modal"
|
|
5
|
+
? new ModalSandboxExecutor(config)
|
|
6
|
+
: new DockerSandboxExecutor(config);
|
|
7
|
+
}
|
|
8
|
+
//# sourceMappingURL=sandbox.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"sandbox.js","sourceRoot":"","sources":["../src/sandbox.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,qBAAqB,EAAE,MAAM,sBAAsB,CAAC;AAC7D,OAAO,EAAE,oBAAoB,EAAE,MAAM,qBAAqB,CAAC;AA4C3D,MAAM,UAAU,qBAAqB,CAAC,MAAoC;IACxE,OAAO,MAAM,CAAC,IAAI,KAAK,OAAO;QAC5B,CAAC,CAAC,IAAI,oBAAoB,CAAC,MAAM,CAAC;QAClC,CAAC,CAAC,IAAI,qBAAqB,CAAC,MAAM,CAAC,CAAC;AACxC,CAAC"}
|
|
@@ -0,0 +1,6 @@
|
|
|
1
|
+
export declare function loadPiSubscriptionAuth(): Promise<string>;
|
|
2
|
+
export declare function assertCodexSubscriptionAuth(value: unknown, source?: string): void;
|
|
3
|
+
export declare function loadCodexSubscriptionAuth(): Promise<string>;
|
|
4
|
+
export declare function subscriptionBearerToken(model: string): Promise<string>;
|
|
5
|
+
export declare function githubToken(): Promise<string | undefined>;
|
|
6
|
+
//# sourceMappingURL=subscription-auth.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"subscription-auth.d.ts","sourceRoot":"","sources":["../src/subscription-auth.ts"],"names":[],"mappings":"AAKA,wBAAsB,sBAAsB,IAAI,OAAO,CAAC,MAAM,CAAC,CAe9D;AAED,wBAAgB,2BAA2B,CAAC,KAAK,EAAE,OAAO,EAAE,MAAM,SAAe,GAAG,IAAI,CAIvF;AAED,wBAAsB,yBAAyB,IAAI,OAAO,CAAC,MAAM,CAAC,CAUjE;AAED,wBAAsB,uBAAuB,CAAC,KAAK,EAAE,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,CAc5E;AAED,wBAAsB,WAAW,IAAI,OAAO,CAAC,MAAM,GAAG,SAAS,CAAC,CAM/D"}
|
|
@@ -0,0 +1,59 @@
|
|
|
1
|
+
import { readFile } from "node:fs/promises";
|
|
2
|
+
import { homedir } from "node:os";
|
|
3
|
+
import { join } from "node:path";
|
|
4
|
+
import { runCommand } from "./process.js";
|
|
5
|
+
export async function loadPiSubscriptionAuth() {
|
|
6
|
+
const raw = process.env.SELFBENCH_PI_AUTH_JSON ??
|
|
7
|
+
(await readFile(join(homedir(), ".pi/agent/auth.json"), "utf8"));
|
|
8
|
+
const parsed = JSON.parse(raw);
|
|
9
|
+
const credential = isRecord(parsed) ? parsed["openai-codex"] : undefined;
|
|
10
|
+
if (!isRecord(credential) ||
|
|
11
|
+
credential.type !== "oauth" ||
|
|
12
|
+
typeof credential.access !== "string" ||
|
|
13
|
+
typeof credential.refresh !== "string") {
|
|
14
|
+
throw new Error("Pi auth does not contain an openai-codex subscription credential");
|
|
15
|
+
}
|
|
16
|
+
return JSON.stringify({ "openai-codex": credential });
|
|
17
|
+
}
|
|
18
|
+
export function assertCodexSubscriptionAuth(value, source = "Codex auth") {
|
|
19
|
+
if (!isRecord(value) || value.auth_mode !== "chatgpt" || !isRecord(value.tokens)) {
|
|
20
|
+
throw new Error(`${source} does not contain a ChatGPT subscription token set`);
|
|
21
|
+
}
|
|
22
|
+
}
|
|
23
|
+
export async function loadCodexSubscriptionAuth() {
|
|
24
|
+
const path = process.env.CODEX_AUTH_JSON_PATH ?? join(homedir(), ".codex/auth.json");
|
|
25
|
+
const parsed = JSON.parse(await readFile(path, "utf8"));
|
|
26
|
+
assertCodexSubscriptionAuth(parsed, path);
|
|
27
|
+
const auth = parsed;
|
|
28
|
+
return JSON.stringify({
|
|
29
|
+
auth_mode: "chatgpt",
|
|
30
|
+
tokens: auth.tokens,
|
|
31
|
+
...(typeof auth.last_refresh === "string" ? { last_refresh: auth.last_refresh } : {}),
|
|
32
|
+
});
|
|
33
|
+
}
|
|
34
|
+
export async function subscriptionBearerToken(model) {
|
|
35
|
+
const result = await runCommand("pi", [
|
|
36
|
+
"auth",
|
|
37
|
+
"print-bearer-token",
|
|
38
|
+
"--provider",
|
|
39
|
+
"openai-codex",
|
|
40
|
+
"--model",
|
|
41
|
+
model,
|
|
42
|
+
]);
|
|
43
|
+
const token = result.stdout.trim();
|
|
44
|
+
if (!token) {
|
|
45
|
+
throw new Error(`Pi returned no openai-codex subscription bearer token for ${model}`);
|
|
46
|
+
}
|
|
47
|
+
return token;
|
|
48
|
+
}
|
|
49
|
+
export async function githubToken() {
|
|
50
|
+
if (process.env.GH_TOKEN) {
|
|
51
|
+
return process.env.GH_TOKEN;
|
|
52
|
+
}
|
|
53
|
+
const result = await runCommand("gh", ["auth", "token"], { allowFailure: true });
|
|
54
|
+
return result.exitCode === 0 && result.stdout.trim() ? result.stdout.trim() : undefined;
|
|
55
|
+
}
|
|
56
|
+
function isRecord(value) {
|
|
57
|
+
return typeof value === "object" && value !== null && !Array.isArray(value);
|
|
58
|
+
}
|
|
59
|
+
//# sourceMappingURL=subscription-auth.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"subscription-auth.js","sourceRoot":"","sources":["../src/subscription-auth.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,EAAE,OAAO,EAAE,MAAM,SAAS,CAAC;AAClC,OAAO,EAAE,IAAI,EAAE,MAAM,WAAW,CAAC;AACjC,OAAO,EAAE,UAAU,EAAE,MAAM,cAAc,CAAC;AAE1C,MAAM,CAAC,KAAK,UAAU,sBAAsB;IAC1C,MAAM,GAAG,GACP,OAAO,CAAC,GAAG,CAAC,sBAAsB;QAClC,CAAC,MAAM,QAAQ,CAAC,IAAI,CAAC,OAAO,EAAE,EAAE,qBAAqB,CAAC,EAAE,MAAM,CAAC,CAAC,CAAC;IACnE,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,CAAY,CAAC;IAC1C,MAAM,UAAU,GAAG,QAAQ,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,cAAc,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;IACzE,IACE,CAAC,QAAQ,CAAC,UAAU,CAAC;QACrB,UAAU,CAAC,IAAI,KAAK,OAAO;QAC3B,OAAO,UAAU,CAAC,MAAM,KAAK,QAAQ;QACrC,OAAO,UAAU,CAAC,OAAO,KAAK,QAAQ,EACtC,CAAC;QACD,MAAM,IAAI,KAAK,CAAC,kEAAkE,CAAC,CAAC;IACtF,CAAC;IACD,OAAO,IAAI,CAAC,SAAS,CAAC,EAAE,cAAc,EAAE,UAAU,EAAE,CAAC,CAAC;AACxD,CAAC;AAED,MAAM,UAAU,2BAA2B,CAAC,KAAc,EAAE,MAAM,GAAG,YAAY;IAC/E,IAAI,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,KAAK,CAAC,SAAS,KAAK,SAAS,IAAI,CAAC,QAAQ,CAAC,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC;QACjF,MAAM,IAAI,KAAK,CAAC,GAAG,MAAM,oDAAoD,CAAC,CAAC;IACjF,CAAC;AACH,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,yBAAyB;IAC7C,MAAM,IAAI,GAAG,OAAO,CAAC,GAAG,CAAC,oBAAoB,IAAI,IAAI,CAAC,OAAO,EAAE,EAAE,kBAAkB,CAAC,CAAC;IACrF,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,MAAM,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC,CAAY,CAAC;IACnE,2BAA2B,CAAC,MAAM,EAAE,IAAI,CAAC,CAAC;IAC1C,MAAM,IAAI,GAAG,MAAiC,CAAC;IAC/C,OAAO,IAAI,CAAC,SAAS,CAAC;QACpB,SAAS,EAAE,SAAS;QACpB,MAAM,EAAE,IAAI,CAAC,MAAM;QACnB,GAAG,CAAC,OAAO,IAAI,CAAC,YAAY,KAAK,QAAQ,CAAC,CAAC,CAAC,EAAE,YAAY,EAAE,IAAI,CAAC,YAAY,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;KACtF,CAAC,CAAC;AACL,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,uBAAuB,CAAC,KAAa;IACzD,MAAM,MAAM,GAAG,MAAM,UAAU,CAAC,IAAI,EAAE;QACpC,MAAM;QACN,oBAAoB;QACpB,YAAY;QACZ,cAAc;QACd,SAAS;QACT,KAAK;KACN,CAAC,CAAC;IACH,MAAM,KAAK,GAAG,MAAM,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC;IACnC,IAAI,CAAC,KAAK,EAAE,CAAC;QACX,MAAM,IAAI,KAAK,CAAC,6DAA6D,KAAK,EAAE,CAAC,CAAC;IACxF,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,WAAW;IAC/B,IAAI,OAAO,CAAC,GAAG,CAAC,QAAQ,EAAE,CAAC;QACzB,OAAO,OAAO,CAAC,GAAG,CAAC,QAAQ,CAAC;IAC9B,CAAC;IACD,MAAM,MAAM,GAAG,MAAM,UAAU,CAAC,IAAI,EAAE,CAAC,MAAM,EAAE,OAAO,CAAC,EAAE,EAAE,YAAY,EAAE,IAAI,EAAE,CAAC,CAAC;IACjF,OAAO,MAAM,CAAC,QAAQ,KAAK,CAAC,IAAI,MAAM,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,SAAS,CAAC;AAC1F,CAAC;AAED,SAAS,QAAQ,CAAC,KAAc;IAC9B,OAAO,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,KAAK,IAAI,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;AAC9E,CAAC"}
|
|
@@ -0,0 +1,6 @@
|
|
|
1
|
+
import { Connection } from "@temporalio/client";
|
|
2
|
+
import { NativeConnection } from "@temporalio/worker";
|
|
3
|
+
import type { SelfBenchConfig } from "./config.js";
|
|
4
|
+
export declare function connectTemporalClient(config: SelfBenchConfig["temporal"]): Promise<Connection>;
|
|
5
|
+
export declare function connectTemporalWorker(config: SelfBenchConfig["temporal"]): Promise<NativeConnection>;
|
|
6
|
+
//# sourceMappingURL=temporal.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"temporal.d.ts","sourceRoot":"","sources":["../src/temporal.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAE,MAAM,oBAAoB,CAAC;AAChD,OAAO,EAAE,gBAAgB,EAAE,MAAM,oBAAoB,CAAC;AACtD,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,aAAa,CAAC;AAEnD,wBAAsB,qBAAqB,CACzC,MAAM,EAAE,eAAe,CAAC,UAAU,CAAC,GAClC,OAAO,CAAC,UAAU,CAAC,CAMrB;AAED,wBAAsB,qBAAqB,CACzC,MAAM,EAAE,eAAe,CAAC,UAAU,CAAC,GAClC,OAAO,CAAC,gBAAgB,CAAC,CAM3B"}
|
package/dist/temporal.js
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
1
|
+
import { Connection } from "@temporalio/client";
|
|
2
|
+
import { NativeConnection } from "@temporalio/worker";
|
|
3
|
+
export async function connectTemporalClient(config) {
|
|
4
|
+
return await Connection.connect({
|
|
5
|
+
address: config.address,
|
|
6
|
+
tls: config.tls,
|
|
7
|
+
...(config.apiKey ? { apiKey: config.apiKey } : {}),
|
|
8
|
+
});
|
|
9
|
+
}
|
|
10
|
+
export async function connectTemporalWorker(config) {
|
|
11
|
+
return await NativeConnection.connect({
|
|
12
|
+
address: config.address,
|
|
13
|
+
tls: config.tls,
|
|
14
|
+
...(config.apiKey ? { apiKey: config.apiKey } : {}),
|
|
15
|
+
});
|
|
16
|
+
}
|
|
17
|
+
//# sourceMappingURL=temporal.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"temporal.js","sourceRoot":"","sources":["../src/temporal.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAE,MAAM,oBAAoB,CAAC;AAChD,OAAO,EAAE,gBAAgB,EAAE,MAAM,oBAAoB,CAAC;AAGtD,MAAM,CAAC,KAAK,UAAU,qBAAqB,CACzC,MAAmC;IAEnC,OAAO,MAAM,UAAU,CAAC,OAAO,CAAC;QAC9B,OAAO,EAAE,MAAM,CAAC,OAAO;QACvB,GAAG,EAAE,MAAM,CAAC,GAAG;QACf,GAAG,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,MAAM,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;KACpD,CAAC,CAAC;AACL,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,qBAAqB,CACzC,MAAmC;IAEnC,OAAO,MAAM,gBAAgB,CAAC,OAAO,CAAC;QACpC,OAAO,EAAE,MAAM,CAAC,OAAO;QACvB,GAAG,EAAE,MAAM,CAAC,GAAG;QACf,GAAG,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,MAAM,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;KACpD,CAAC,CAAC;AACL,CAAC"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"validate-main.d.ts","sourceRoot":"","sources":["../src/validate-main.ts"],"names":[],"mappings":""}
|
|
@@ -0,0 +1,137 @@
|
|
|
1
|
+
#!/usr/bin/env node
|
|
2
|
+
import { access, mkdir, readdir, writeFile } from "node:fs/promises";
|
|
3
|
+
import { basename, join, resolve } from "node:path";
|
|
4
|
+
import { parseArgs } from "node:util";
|
|
5
|
+
import { harborInfrastructureError, readHarborJobResult } from "./harbor-results.js";
|
|
6
|
+
import { parallelMap } from "./parallel.js";
|
|
7
|
+
import { runCommand } from "./process.js";
|
|
8
|
+
const parsed = parseArgs({
|
|
9
|
+
options: {
|
|
10
|
+
tasks: { type: "string" },
|
|
11
|
+
jobs: { type: "string" },
|
|
12
|
+
environment: { type: "string", default: "modal" },
|
|
13
|
+
concurrency: { type: "string", default: "10" },
|
|
14
|
+
help: { type: "boolean", short: "h" },
|
|
15
|
+
},
|
|
16
|
+
strict: true,
|
|
17
|
+
});
|
|
18
|
+
if (parsed.values.help) {
|
|
19
|
+
console.log(`Run Harbor nop and oracle gates for expanded SelfBench tasks.
|
|
20
|
+
|
|
21
|
+
Usage:
|
|
22
|
+
selfbench-validate --tasks DIRECTORY --jobs DIRECTORY [options]
|
|
23
|
+
|
|
24
|
+
Options:
|
|
25
|
+
--environment NAME Harbor environment (default: modal)
|
|
26
|
+
--concurrency N Concurrent tasks (default: 10)
|
|
27
|
+
-h, --help Show this help`);
|
|
28
|
+
process.exit(0);
|
|
29
|
+
}
|
|
30
|
+
const tasksDirectory = resolve(parsed.values.tasks ?? fail("--tasks is required"));
|
|
31
|
+
const jobsDirectory = resolve(parsed.values.jobs ?? fail("--jobs is required"));
|
|
32
|
+
const environment = parsed.values.environment ?? "modal";
|
|
33
|
+
if (environment !== "modal" && environment !== "docker") {
|
|
34
|
+
throw new Error("--environment must be modal or docker");
|
|
35
|
+
}
|
|
36
|
+
const concurrency = positiveInteger(parsed.values.concurrency, "--concurrency");
|
|
37
|
+
const directories = (await Promise.all((await readdir(tasksDirectory)).map(async (name) => {
|
|
38
|
+
const directory = join(tasksDirectory, name);
|
|
39
|
+
return await access(join(directory, "task.toml")).then(() => directory, () => undefined);
|
|
40
|
+
}))).filter((directory) => directory !== undefined);
|
|
41
|
+
await mkdir(jobsDirectory, { recursive: true });
|
|
42
|
+
const tasks = await parallelMap(directories.sort(), concurrency, async (taskDirectory) => {
|
|
43
|
+
const taskId = basename(taskDirectory);
|
|
44
|
+
console.error(`validating ${taskId}`);
|
|
45
|
+
try {
|
|
46
|
+
const nop = await runGate(taskDirectory, taskId, "nop", environment, jobsDirectory);
|
|
47
|
+
const oracle = await runGate(taskDirectory, taskId, "oracle", environment, jobsDirectory);
|
|
48
|
+
const accepted = nop.passed && oracle.passed;
|
|
49
|
+
console.error(`${taskId}: ${accepted ? "accepted" : "rejected"}`);
|
|
50
|
+
return { taskId, status: "validated", accepted, nop, oracle };
|
|
51
|
+
}
|
|
52
|
+
catch (error) {
|
|
53
|
+
const message = error instanceof Error ? error.message : String(error);
|
|
54
|
+
console.error(`${taskId}: error: ${message}`);
|
|
55
|
+
return { taskId, status: "error", accepted: false, error: message };
|
|
56
|
+
}
|
|
57
|
+
});
|
|
58
|
+
const report = {
|
|
59
|
+
schemaVersion: 1,
|
|
60
|
+
generatedAt: new Date().toISOString(),
|
|
61
|
+
sourceTasks: tasksDirectory,
|
|
62
|
+
environment,
|
|
63
|
+
summary: {
|
|
64
|
+
total: tasks.length,
|
|
65
|
+
accepted: tasks.filter((task) => task.accepted).length,
|
|
66
|
+
rejected: tasks.filter((task) => task.status === "validated" && !task.accepted).length,
|
|
67
|
+
errors: tasks.filter((task) => task.status === "error").length,
|
|
68
|
+
},
|
|
69
|
+
tasks,
|
|
70
|
+
};
|
|
71
|
+
await writeFile(join(jobsDirectory, "validation-summary.json"), `${JSON.stringify(report, null, 2)}\n`);
|
|
72
|
+
console.log(JSON.stringify(report.summary, null, 2));
|
|
73
|
+
async function runGate(taskDirectory, taskId, agent, environment, jobsDirectory) {
|
|
74
|
+
const jobName = `${taskId}-${agent}-${crypto.randomUUID().slice(0, 8)}`;
|
|
75
|
+
const process = await runCommand("harbor", [
|
|
76
|
+
"run",
|
|
77
|
+
"--path",
|
|
78
|
+
taskDirectory,
|
|
79
|
+
"--agent",
|
|
80
|
+
agent,
|
|
81
|
+
"--env",
|
|
82
|
+
environment,
|
|
83
|
+
"--job-name",
|
|
84
|
+
jobName,
|
|
85
|
+
"--jobs-dir",
|
|
86
|
+
jobsDirectory,
|
|
87
|
+
"--n-concurrent",
|
|
88
|
+
"1",
|
|
89
|
+
"--max-retries",
|
|
90
|
+
"1",
|
|
91
|
+
"--delete",
|
|
92
|
+
"--yes",
|
|
93
|
+
"--quiet",
|
|
94
|
+
], { allowFailure: true, timeoutMs: 4 * 60 * 60 * 1000 });
|
|
95
|
+
if (process.exitCode !== 0) {
|
|
96
|
+
throw new Error(`Harbor ${agent} exited ${process.exitCode}: ${process.stderr.slice(-1_000)}`);
|
|
97
|
+
}
|
|
98
|
+
const result = await readHarborJobResult(jobsDirectory, jobName);
|
|
99
|
+
const infrastructure = harborInfrastructureError(result.trial);
|
|
100
|
+
if (infrastructure) {
|
|
101
|
+
throw new Error(`Harbor ${agent} infrastructure failure: ${infrastructure}`);
|
|
102
|
+
}
|
|
103
|
+
const rewards = rewardValues(result.trial);
|
|
104
|
+
const passed = agent === "nop"
|
|
105
|
+
? rewards.fail_to_pass === 0 &&
|
|
106
|
+
(rewards.pass_to_pass ?? 0) >= 1 &&
|
|
107
|
+
(rewards.setup_completed ?? 0) >= 1
|
|
108
|
+
: (rewards.patch_applied ?? 0) >= 1 &&
|
|
109
|
+
(rewards.fail_to_pass ?? 0) >= 1 &&
|
|
110
|
+
(rewards.pass_to_pass ?? 0) >= 1 &&
|
|
111
|
+
(rewards.deterministic ?? 0) >= 1 &&
|
|
112
|
+
(rewards.setup_completed ?? 0) >= 1;
|
|
113
|
+
return { passed, rewards };
|
|
114
|
+
}
|
|
115
|
+
function rewardValues(value) {
|
|
116
|
+
if (!isRecord(value) || !isRecord(value.verifier_result)) {
|
|
117
|
+
return {};
|
|
118
|
+
}
|
|
119
|
+
const raw = value.verifier_result.rewards;
|
|
120
|
+
return isRecord(raw)
|
|
121
|
+
? Object.fromEntries(Object.entries(raw).filter((entry) => typeof entry[1] === "number"))
|
|
122
|
+
: {};
|
|
123
|
+
}
|
|
124
|
+
function positiveInteger(value, label) {
|
|
125
|
+
const number = Number(value);
|
|
126
|
+
if (!Number.isInteger(number) || number < 1) {
|
|
127
|
+
throw new Error(`${label} must be a positive integer`);
|
|
128
|
+
}
|
|
129
|
+
return number;
|
|
130
|
+
}
|
|
131
|
+
function isRecord(value) {
|
|
132
|
+
return typeof value === "object" && value !== null && !Array.isArray(value);
|
|
133
|
+
}
|
|
134
|
+
function fail(message) {
|
|
135
|
+
throw new Error(message);
|
|
136
|
+
}
|
|
137
|
+
//# sourceMappingURL=validate-main.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"validate-main.js","sourceRoot":"","sources":["../src/validate-main.ts"],"names":[],"mappings":";AAEA,OAAO,EAAE,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,SAAS,EAAE,MAAM,kBAAkB,CAAC;AACrE,OAAO,EAAE,QAAQ,EAAE,IAAI,EAAE,OAAO,EAAE,MAAM,WAAW,CAAC;AACpD,OAAO,EAAE,SAAS,EAAE,MAAM,WAAW,CAAC;AACtC,OAAO,EAAE,yBAAyB,EAAE,mBAAmB,EAAE,MAAM,qBAAqB,CAAC;AACrF,OAAO,EAAE,WAAW,EAAE,MAAM,eAAe,CAAC;AAC5C,OAAO,EAAE,UAAU,EAAE,MAAM,cAAc,CAAC;AAE1C,MAAM,MAAM,GAAG,SAAS,CAAC;IACvB,OAAO,EAAE;QACP,KAAK,EAAE,EAAE,IAAI,EAAE,QAAQ,EAAE;QACzB,IAAI,EAAE,EAAE,IAAI,EAAE,QAAQ,EAAE;QACxB,WAAW,EAAE,EAAE,IAAI,EAAE,QAAQ,EAAE,OAAO,EAAE,OAAO,EAAE;QACjD,WAAW,EAAE,EAAE,IAAI,EAAE,QAAQ,EAAE,OAAO,EAAE,IAAI,EAAE;QAC9C,IAAI,EAAE,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,GAAG,EAAE;KACtC;IACD,MAAM,EAAE,IAAI;CACb,CAAC,CAAC;AACH,IAAI,MAAM,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC;IACvB,OAAO,CAAC,GAAG,CAAC;;;;;;;;qCAQuB,CAAC,CAAC;IACrC,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;AAClB,CAAC;AAED,MAAM,cAAc,GAAG,OAAO,CAAC,MAAM,CAAC,MAAM,CAAC,KAAK,IAAI,IAAI,CAAC,qBAAqB,CAAC,CAAC,CAAC;AACnF,MAAM,aAAa,GAAG,OAAO,CAAC,MAAM,CAAC,MAAM,CAAC,IAAI,IAAI,IAAI,CAAC,oBAAoB,CAAC,CAAC,CAAC;AAChF,MAAM,WAAW,GAAG,MAAM,CAAC,MAAM,CAAC,WAAW,IAAI,OAAO,CAAC;AACzD,IAAI,WAAW,KAAK,OAAO,IAAI,WAAW,KAAK,QAAQ,EAAE,CAAC;IACxD,MAAM,IAAI,KAAK,CAAC,uCAAuC,CAAC,CAAC;AAC3D,CAAC;AACD,MAAM,WAAW,GAAG,eAAe,CAAC,MAAM,CAAC,MAAM,CAAC,WAAW,EAAE,eAAe,CAAC,CAAC;AAChF,MAAM,WAAW,GAAG,CAClB,MAAM,OAAO,CAAC,GAAG,CACf,CACE,MAAM,OAAO,CAAC,cAAc,CAAC,CAC9B,CAAC,GAAG,CAAC,KAAK,EAAE,IAAI,EAAE,EAAE;IACnB,MAAM,SAAS,GAAG,IAAI,CAAC,cAAc,EAAE,IAAI,CAAC,CAAC;IAC7C,OAAO,MAAM,MAAM,CAAC,IAAI,CAAC,SAAS,EAAE,WAAW,CAAC,CAAC,CAAC,IAAI,CACpD,GAAG,EAAE,CAAC,SAAS,EACf,GAAG,EAAE,CAAC,SAAS,CAChB,CAAC;AACJ,CAAC,CAAC,CACH,CACF,CAAC,MAAM,CAAC,CAAC,SAAS,EAAuB,EAAE,CAAC,SAAS,KAAK,SAAS,CAAC,CAAC;AACtE,MAAM,KAAK,CAAC,aAAa,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;AAEhD,MAAM,KAAK,GAAG,MAAM,WAAW,CAAC,WAAW,CAAC,IAAI,EAAE,EAAE,WAAW,EAAE,KAAK,EAAE,aAAa,EAAE,EAAE;IACvF,MAAM,MAAM,GAAG,QAAQ,CAAC,aAAa,CAAC,CAAC;IACvC,OAAO,CAAC,KAAK,CAAC,cAAc,MAAM,EAAE,CAAC,CAAC;IACtC,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,MAAM,OAAO,CAAC,aAAa,EAAE,MAAM,EAAE,KAAK,EAAE,WAAW,EAAE,aAAa,CAAC,CAAC;QACpF,MAAM,MAAM,GAAG,MAAM,OAAO,CAAC,aAAa,EAAE,MAAM,EAAE,QAAQ,EAAE,WAAW,EAAE,aAAa,CAAC,CAAC;QAC1F,MAAM,QAAQ,GAAG,GAAG,CAAC,MAAM,IAAI,MAAM,CAAC,MAAM,CAAC;QAC7C,OAAO,CAAC,KAAK,CAAC,GAAG,MAAM,KAAK,QAAQ,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,CAAC,UAAU,EAAE,CAAC,CAAC;QAClE,OAAO,EAAE,MAAM,EAAE,MAAM,EAAE,WAAoB,EAAE,QAAQ,EAAE,GAAG,EAAE,MAAM,EAAE,CAAC;IACzE,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,MAAM,OAAO,GAAG,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;QACvE,OAAO,CAAC,KAAK,CAAC,GAAG,MAAM,YAAY,OAAO,EAAE,CAAC,CAAC;QAC9C,OAAO,EAAE,MAAM,EAAE,MAAM,EAAE,OAAgB,EAAE,QAAQ,EAAE,KAAK,EAAE,KAAK,EAAE,OAAO,EAAE,CAAC;IAC/E,CAAC;AACH,CAAC,CAAC,CAAC;AACH,MAAM,MAAM,GAAG;IACb,aAAa,EAAE,CAAC;IAChB,WAAW,EAAE,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE;IACrC,WAAW,EAAE,cAAc;IAC3B,WAAW;IACX,OAAO,EAAE;QACP,KAAK,EAAE,KAAK,CAAC,MAAM;QACnB,QAAQ,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,MAAM;QACtD,QAAQ,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,MAAM,KAAK,WAAW,IAAI,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,MAAM;QACtF,MAAM,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,MAAM,KAAK,OAAO,CAAC,CAAC,MAAM;KAC/D;IACD,KAAK;CACN,CAAC;AACF,MAAM,SAAS,CACb,IAAI,CAAC,aAAa,EAAE,yBAAyB,CAAC,EAC9C,GAAG,IAAI,CAAC,SAAS,CAAC,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC,IAAI,CACvC,CAAC;AACF,OAAO,CAAC,GAAG,CAAC,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,OAAO,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC;AAErD,KAAK,UAAU,OAAO,CACpB,aAAqB,EACrB,MAAc,EACd,KAAuB,EACvB,WAA+B,EAC/B,aAAqB;IAErB,MAAM,OAAO,GAAG,GAAG,MAAM,IAAI,KAAK,IAAI,MAAM,CAAC,UAAU,EAAE,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,EAAE,CAAC;IACxE,MAAM,OAAO,GAAG,MAAM,UAAU,CAC9B,QAAQ,EACR;QACE,KAAK;QACL,QAAQ;QACR,aAAa;QACb,SAAS;QACT,KAAK;QACL,OAAO;QACP,WAAW;QACX,YAAY;QACZ,OAAO;QACP,YAAY;QACZ,aAAa;QACb,gBAAgB;QAChB,GAAG;QACH,eAAe;QACf,GAAG;QACH,UAAU;QACV,OAAO;QACP,SAAS;KACV,EACD,EAAE,YAAY,EAAE,IAAI,EAAE,SAAS,EAAE,CAAC,GAAG,EAAE,GAAG,EAAE,GAAG,IAAI,EAAE,CACtD,CAAC;IACF,IAAI,OAAO,CAAC,QAAQ,KAAK,CAAC,EAAE,CAAC;QAC3B,MAAM,IAAI,KAAK,CAAC,UAAU,KAAK,WAAW,OAAO,CAAC,QAAQ,KAAK,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC;IACjG,CAAC;IACD,MAAM,MAAM,GAAG,MAAM,mBAAmB,CAAC,aAAa,EAAE,OAAO,CAAC,CAAC;IACjE,MAAM,cAAc,GAAG,yBAAyB,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;IAC/D,IAAI,cAAc,EAAE,CAAC;QACnB,MAAM,IAAI,KAAK,CAAC,UAAU,KAAK,4BAA4B,cAAc,EAAE,CAAC,CAAC;IAC/E,CAAC;IACD,MAAM,OAAO,GAAG,YAAY,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;IAC3C,MAAM,MAAM,GACV,KAAK,KAAK,KAAK;QACb,CAAC,CAAC,OAAO,CAAC,YAAY,KAAK,CAAC;YAC1B,CAAC,OAAO,CAAC,YAAY,IAAI,CAAC,CAAC,IAAI,CAAC;YAChC,CAAC,OAAO,CAAC,eAAe,IAAI,CAAC,CAAC,IAAI,CAAC;QACrC,CAAC,CAAC,CAAC,OAAO,CAAC,aAAa,IAAI,CAAC,CAAC,IAAI,CAAC;YACjC,CAAC,OAAO,CAAC,YAAY,IAAI,CAAC,CAAC,IAAI,CAAC;YAChC,CAAC,OAAO,CAAC,YAAY,IAAI,CAAC,CAAC,IAAI,CAAC;YAChC,CAAC,OAAO,CAAC,aAAa,IAAI,CAAC,CAAC,IAAI,CAAC;YACjC,CAAC,OAAO,CAAC,eAAe,IAAI,CAAC,CAAC,IAAI,CAAC,CAAC;IAC1C,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,CAAC;AAC7B,CAAC;AAED,SAAS,YAAY,CAAC,KAAc;IAClC,IAAI,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,CAAC,QAAQ,CAAC,KAAK,CAAC,eAAe,CAAC,EAAE,CAAC;QACzD,OAAO,EAAE,CAAC;IACZ,CAAC;IACD,MAAM,GAAG,GAAG,KAAK,CAAC,eAAe,CAAC,OAAO,CAAC;IAC1C,OAAO,QAAQ,CAAC,GAAG,CAAC;QAClB,CAAC,CAAC,MAAM,CAAC,WAAW,CAChB,MAAM,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,MAAM,CACxB,CAAC,KAAK,EAA6B,EAAE,CAAC,OAAO,KAAK,CAAC,CAAC,CAAC,KAAK,QAAQ,CACnE,CACF;QACH,CAAC,CAAC,EAAE,CAAC;AACT,CAAC;AAED,SAAS,eAAe,CAAC,KAAyB,EAAE,KAAa;IAC/D,MAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,CAAC;IAC7B,IAAI,CAAC,MAAM,CAAC,SAAS,CAAC,MAAM,CAAC,IAAI,MAAM,GAAG,CAAC,EAAE,CAAC;QAC5C,MAAM,IAAI,KAAK,CAAC,GAAG,KAAK,6BAA6B,CAAC,CAAC;IACzD,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,SAAS,QAAQ,CAAC,KAAc;IAC9B,OAAO,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,KAAK,IAAI,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;AAC9E,CAAC;AAED,SAAS,IAAI,CAAC,OAAe;IAC3B,MAAM,IAAI,KAAK,CAAC,OAAO,CAAC,CAAC;AAC3B,CAAC"}
|
|
@@ -0,0 +1,10 @@
|
|
|
1
|
+
import type { TaskDefinition } from "./contracts.js";
|
|
2
|
+
export declare function validationRepairPaths(originalTestPatch: string): readonly string[];
|
|
3
|
+
export declare function assertValidationRepair(original: TaskDefinition, repaired: TaskDefinition, originalTestPatch: string, changedPaths: readonly string[]): void;
|
|
4
|
+
export declare function validationRepairPrompt(input: {
|
|
5
|
+
readonly definition: TaskDefinition;
|
|
6
|
+
readonly authenticRequest: string;
|
|
7
|
+
readonly diagnostics: string;
|
|
8
|
+
readonly allowedPaths: readonly string[];
|
|
9
|
+
}): string;
|
|
10
|
+
//# sourceMappingURL=validation-repair.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"validation-repair.d.ts","sourceRoot":"","sources":["../src/validation-repair.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,gBAAgB,CAAC;AAGrD,wBAAgB,qBAAqB,CAAC,iBAAiB,EAAE,MAAM,GAAG,SAAS,MAAM,EAAE,CAUlF;AAcD,wBAAgB,sBAAsB,CACpC,QAAQ,EAAE,cAAc,EACxB,QAAQ,EAAE,cAAc,EACxB,iBAAiB,EAAE,MAAM,EACzB,YAAY,EAAE,SAAS,MAAM,EAAE,GAC9B,IAAI,CAwBN;AAED,wBAAgB,sBAAsB,CAAC,KAAK,EAAE;IAC5C,QAAQ,CAAC,UAAU,EAAE,cAAc,CAAC;IACpC,QAAQ,CAAC,gBAAgB,EAAE,MAAM,CAAC;IAClC,QAAQ,CAAC,WAAW,EAAE,MAAM,CAAC;IAC7B,QAAQ,CAAC,YAAY,EAAE,SAAS,MAAM,EAAE,CAAC;CAC1C,GAAG,MAAM,CA8BT"}
|
|
@@ -0,0 +1,72 @@
|
|
|
1
|
+
import { patchPaths } from "./repair.js";
|
|
2
|
+
export function validationRepairPaths(originalTestPatch) {
|
|
3
|
+
const paths = patchPaths(originalTestPatch);
|
|
4
|
+
if (paths.length === 0) {
|
|
5
|
+
throw new Error("original held-out test patch changes no files");
|
|
6
|
+
}
|
|
7
|
+
const nonTestPaths = paths.filter((path) => !isTestOnlyPath(path));
|
|
8
|
+
if (nonTestPaths.length > 0) {
|
|
9
|
+
throw new Error(`validation repair requires test-only patch paths: ${nonTestPaths.join(", ")}`);
|
|
10
|
+
}
|
|
11
|
+
return paths;
|
|
12
|
+
}
|
|
13
|
+
function isTestOnlyPath(path) {
|
|
14
|
+
const segments = path.toLowerCase().split("/");
|
|
15
|
+
const filename = segments.at(-1) ?? "";
|
|
16
|
+
return (segments.some((segment) => /^(?:tests?|__tests__|e2e|integration|fixtures?|selfbench(?:-tests|_tests)?|\.selfbench-tests)$/.test(segment)) || /(?:^|\.)(?:test|spec)\.[^.]+$/.test(filename));
|
|
17
|
+
}
|
|
18
|
+
export function assertValidationRepair(original, repaired, originalTestPatch, changedPaths) {
|
|
19
|
+
const immutableKeys = [
|
|
20
|
+
"schemaVersion",
|
|
21
|
+
"difficulty",
|
|
22
|
+
"taskId",
|
|
23
|
+
"repo",
|
|
24
|
+
"baseCommit",
|
|
25
|
+
"workdir",
|
|
26
|
+
"sourcePr",
|
|
27
|
+
"sourceUrl",
|
|
28
|
+
"prompt",
|
|
29
|
+
];
|
|
30
|
+
for (const key of immutableKeys) {
|
|
31
|
+
if (JSON.stringify(repaired[key]) !== JSON.stringify(original[key])) {
|
|
32
|
+
throw new Error(`validation repair changed immutable definition field ${key}`);
|
|
33
|
+
}
|
|
34
|
+
}
|
|
35
|
+
const allowed = new Set(validationRepairPaths(originalTestPatch));
|
|
36
|
+
const outside = changedPaths.filter((path) => !allowed.has(path));
|
|
37
|
+
if (outside.length > 0) {
|
|
38
|
+
throw new Error(`validation repair changed files outside held-out tests: ${outside.join(", ")}`);
|
|
39
|
+
}
|
|
40
|
+
}
|
|
41
|
+
export function validationRepairPrompt(input) {
|
|
42
|
+
return `Repair the verifier harness and held-out tests for SelfBench task ${input.definition.taskId}.
|
|
43
|
+
|
|
44
|
+
The repository is the exact base snapshot with the current held-out test patch already applied. You may edit only these existing held-out test files:
|
|
45
|
+
|
|
46
|
+
${input.allowedPaths.map((path) => `- ${path}`).join("\n")}
|
|
47
|
+
|
|
48
|
+
The authentic engineer request is:
|
|
49
|
+
|
|
50
|
+
<authentic_request>
|
|
51
|
+
${input.authenticRequest.trim()}
|
|
52
|
+
</authentic_request>
|
|
53
|
+
|
|
54
|
+
The failed nop/oracle validation diagnostics are:
|
|
55
|
+
|
|
56
|
+
<validation_diagnostics>
|
|
57
|
+
${input.diagnostics.trim()}
|
|
58
|
+
</validation_diagnostics>
|
|
59
|
+
|
|
60
|
+
Fix the task so its repository-native verifier is rigorous and runnable. You may edit the allowed held-out tests and /work/definition.json. In definition.json you may change only setupCommand, testCommand, failToPass, passToPass, testPaths, toolchains, timeouts, and resources. Never change task identity, difficulty, repository, base commit, workdir, source pull request, prompt, or reference solution.
|
|
61
|
+
|
|
62
|
+
The required validation split is:
|
|
63
|
+
- nop: the base repository plus held-out tests must make failToPass fail while passToPass succeeds;
|
|
64
|
+
- oracle: after /work/task/solution/gold.patch is applied, failToPass and passToPass must both succeed deterministically.
|
|
65
|
+
|
|
66
|
+
Use the verifier diagnostics rather than guessing. Inspect repository package scripts and CI for the native test command. Keep {tests} as a list-safe placeholder; do not quote the whole placeholder, assign it to one scalar, or hard-code selected paths elsewhere. Use one test mode/bundler per command instead of chaining equivalent suites. Put dependency installation, native builds, fixture generation, and browser/runtime installation in setupCommand, not testCommand. Pin the repository's declared package manager and use frozen installs. Browser-backed tests must install required browser binaries and OS dependencies during setup. Run focused nop/oracle-style commands in the sandbox when feasible.
|
|
67
|
+
|
|
68
|
+
Do not weaken, delete, or skip requested behavioral coverage merely to make commands exit zero. A passing base with no solution is not acceptable. If no rigorous runnable split exists, leave the files and definition unchanged and explain why in the final response.
|
|
69
|
+
|
|
70
|
+
Inspect /work/gold.patch and /work/test.sh only to verify the intended behavior and oracle seam, never to couple assertions to private implementation details. Do not modify application code, the instruction, gold patch, or commit files. Finish with the repaired definition at /work/definition.json and any held-out test edits present in the working tree; do not commit.`;
|
|
71
|
+
}
|
|
72
|
+
//# sourceMappingURL=validation-repair.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"validation-repair.js","sourceRoot":"","sources":["../src/validation-repair.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AAEzC,MAAM,UAAU,qBAAqB,CAAC,iBAAyB;IAC7D,MAAM,KAAK,GAAG,UAAU,CAAC,iBAAiB,CAAC,CAAC;IAC5C,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACvB,MAAM,IAAI,KAAK,CAAC,+CAA+C,CAAC,CAAC;IACnE,CAAC;IACD,MAAM,YAAY,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC,cAAc,CAAC,IAAI,CAAC,CAAC,CAAC;IACnE,IAAI,YAAY,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAC5B,MAAM,IAAI,KAAK,CAAC,qDAAqD,YAAY,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAClG,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC;AAED,SAAS,cAAc,CAAC,IAAY;IAClC,MAAM,QAAQ,GAAG,IAAI,CAAC,WAAW,EAAE,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;IAC/C,MAAM,QAAQ,GAAG,QAAQ,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;IACvC,OAAO,CACL,QAAQ,CAAC,IAAI,CAAC,CAAC,OAAO,EAAE,EAAE,CACxB,gGAAgG,CAAC,IAAI,CACnG,OAAO,CACR,CACF,IAAI,+BAA+B,CAAC,IAAI,CAAC,QAAQ,CAAC,CACpD,CAAC;AACJ,CAAC;AAED,MAAM,UAAU,sBAAsB,CACpC,QAAwB,EACxB,QAAwB,EACxB,iBAAyB,EACzB,YAA+B;IAE/B,MAAM,aAAa,GAAG;QACpB,eAAe;QACf,YAAY;QACZ,QAAQ;QACR,MAAM;QACN,YAAY;QACZ,SAAS;QACT,UAAU;QACV,WAAW;QACX,QAAQ;KACA,CAAC;IACX,KAAK,MAAM,GAAG,IAAI,aAAa,EAAE,CAAC;QAChC,IAAI,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,KAAK,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC;YACpE,MAAM,IAAI,KAAK,CAAC,wDAAwD,GAAG,EAAE,CAAC,CAAC;QACjF,CAAC;IACH,CAAC;IACD,MAAM,OAAO,GAAG,IAAI,GAAG,CAAC,qBAAqB,CAAC,iBAAiB,CAAC,CAAC,CAAC;IAClE,MAAM,OAAO,GAAG,YAAY,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC;IAClE,IAAI,OAAO,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QACvB,MAAM,IAAI,KAAK,CACb,2DAA2D,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAChF,CAAC;IACJ,CAAC;AACH,CAAC;AAED,MAAM,UAAU,sBAAsB,CAAC,KAKtC;IACC,OAAO,qEAAqE,KAAK,CAAC,UAAU,CAAC,MAAM;;;;EAInG,KAAK,CAAC,YAAY,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,KAAK,IAAI,EAAE,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC;;;;;EAKxD,KAAK,CAAC,gBAAgB,CAAC,IAAI,EAAE;;;;;;EAM7B,KAAK,CAAC,WAAW,CAAC,IAAI,EAAE;;;;;;;;;;;;;kXAawV,CAAC;AACnX,CAAC"}
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"worker-main.d.ts","sourceRoot":"","sources":["../src/worker-main.ts"],"names":[],"mappings":""}
|
|
@@ -0,0 +1,30 @@
|
|
|
1
|
+
import { existsSync } from "node:fs";
|
|
2
|
+
import { fileURLToPath } from "node:url";
|
|
3
|
+
import { Worker } from "@temporalio/worker";
|
|
4
|
+
import { createActivities } from "./activities.js";
|
|
5
|
+
import { loadConfig } from "./config.js";
|
|
6
|
+
import { removeEmptyModalCredentialOverrides } from "./modal-auth.js";
|
|
7
|
+
import { runCommand } from "./process.js";
|
|
8
|
+
import { connectTemporalWorker } from "./temporal.js";
|
|
9
|
+
removeEmptyModalCredentialOverrides();
|
|
10
|
+
const config = loadConfig();
|
|
11
|
+
if (config.execution.kind === "docker" || config.harborEnvironment === "docker") {
|
|
12
|
+
await runCommand("docker", ["info"], { timeoutMs: 30_000 });
|
|
13
|
+
await runCommand("docker", ["compose", "version"], { timeoutMs: 30_000 });
|
|
14
|
+
}
|
|
15
|
+
const connection = await connectTemporalWorker(config.temporal);
|
|
16
|
+
const javascriptWorkflow = fileURLToPath(new URL("./workflow.js", import.meta.url));
|
|
17
|
+
const workflowsPath = existsSync(javascriptWorkflow)
|
|
18
|
+
? javascriptWorkflow
|
|
19
|
+
: fileURLToPath(new URL("./workflow.ts", import.meta.url));
|
|
20
|
+
const worker = await Worker.create({
|
|
21
|
+
connection,
|
|
22
|
+
namespace: config.temporal.namespace,
|
|
23
|
+
taskQueue: config.temporal.taskQueue,
|
|
24
|
+
workflowsPath,
|
|
25
|
+
activities: createActivities(config),
|
|
26
|
+
maxConcurrentActivityTaskExecutions: config.activityConcurrency,
|
|
27
|
+
});
|
|
28
|
+
console.log(`SelfBench worker polling ${config.temporal.namespace}/${config.temporal.taskQueue} with activity concurrency ${config.activityConcurrency}`);
|
|
29
|
+
await worker.run();
|
|
30
|
+
//# sourceMappingURL=worker-main.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"worker-main.js","sourceRoot":"","sources":["../src/worker-main.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAE,MAAM,SAAS,CAAC;AACrC,OAAO,EAAE,aAAa,EAAE,MAAM,UAAU,CAAC;AACzC,OAAO,EAAE,MAAM,EAAE,MAAM,oBAAoB,CAAC;AAC5C,OAAO,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AACnD,OAAO,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AACzC,OAAO,EAAE,mCAAmC,EAAE,MAAM,iBAAiB,CAAC;AACtE,OAAO,EAAE,UAAU,EAAE,MAAM,cAAc,CAAC;AAC1C,OAAO,EAAE,qBAAqB,EAAE,MAAM,eAAe,CAAC;AAEtD,mCAAmC,EAAE,CAAC;AACtC,MAAM,MAAM,GAAG,UAAU,EAAE,CAAC;AAC5B,IAAI,MAAM,CAAC,SAAS,CAAC,IAAI,KAAK,QAAQ,IAAI,MAAM,CAAC,iBAAiB,KAAK,QAAQ,EAAE,CAAC;IAChF,MAAM,UAAU,CAAC,QAAQ,EAAE,CAAC,MAAM,CAAC,EAAE,EAAE,SAAS,EAAE,MAAM,EAAE,CAAC,CAAC;IAC5D,MAAM,UAAU,CAAC,QAAQ,EAAE,CAAC,SAAS,EAAE,SAAS,CAAC,EAAE,EAAE,SAAS,EAAE,MAAM,EAAE,CAAC,CAAC;AAC5E,CAAC;AACD,MAAM,UAAU,GAAG,MAAM,qBAAqB,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC;AAChE,MAAM,kBAAkB,GAAG,aAAa,CAAC,IAAI,GAAG,CAAC,eAAe,EAAE,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;AACpF,MAAM,aAAa,GAAG,UAAU,CAAC,kBAAkB,CAAC;IAClD,CAAC,CAAC,kBAAkB;IACpB,CAAC,CAAC,aAAa,CAAC,IAAI,GAAG,CAAC,eAAe,EAAE,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;AAC7D,MAAM,MAAM,GAAG,MAAM,MAAM,CAAC,MAAM,CAAC;IACjC,UAAU;IACV,SAAS,EAAE,MAAM,CAAC,QAAQ,CAAC,SAAS;IACpC,SAAS,EAAE,MAAM,CAAC,QAAQ,CAAC,SAAS;IACpC,aAAa;IACb,UAAU,EAAE,gBAAgB,CAAC,MAAM,CAAC;IACpC,mCAAmC,EAAE,MAAM,CAAC,mBAAmB;CAChE,CAAC,CAAC;AACH,OAAO,CAAC,GAAG,CACT,4BAA4B,MAAM,CAAC,QAAQ,CAAC,SAAS,IAAI,MAAM,CAAC,QAAQ,CAAC,SAAS,8BAA8B,MAAM,CAAC,mBAAmB,EAAE,CAC7I,CAAC;AACF,MAAM,MAAM,CAAC,GAAG,EAAE,CAAC"}
|