@hona/openeval 0.2.0
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/JUDGING.md +123 -0
- package/LICENSE +21 -0
- package/README.md +133 -0
- package/package.json +39 -0
- package/src/app/add-models.ts +18 -0
- package/src/app/context.ts +15 -0
- package/src/app/cost-plan.ts +118 -0
- package/src/app/eval-state.ts +10 -0
- package/src/app/execution-budget.ts +65 -0
- package/src/app/execution-runtime.ts +63 -0
- package/src/app/input-fingerprints.ts +67 -0
- package/src/app/judge-evidence.ts +71 -0
- package/src/app/judge-run.ts +89 -0
- package/src/app/load-benchmark.ts +267 -0
- package/src/app/merge-runs.ts +117 -0
- package/src/app/monitor-policy.ts +49 -0
- package/src/app/plan-benchmark.ts +127 -0
- package/src/app/read-results.ts +524 -0
- package/src/app/read-run.ts +43 -0
- package/src/app/rejudge.ts +127 -0
- package/src/app/retry-run.ts +87 -0
- package/src/app/run-benchmark.ts +401 -0
- package/src/app/run-eval-pipeline.ts +333 -0
- package/src/app/run-eval.ts +93 -0
- package/src/app/scores.ts +54 -0
- package/src/app/serve-results.ts +252 -0
- package/src/app/snapshot-run.ts +42 -0
- package/src/app/watch-evidence.ts +61 -0
- package/src/cli.ts +145 -0
- package/src/evidence.ts +40 -0
- package/src/index.ts +53 -0
- package/src/infra/containers/candidate.ts +218 -0
- package/src/infra/containers/docker.ts +31 -0
- package/src/infra/containers/oci.ts +335 -0
- package/src/infra/containers/runtime/.dockerignore +7 -0
- package/src/infra/containers/runtime/Dockerfile +19 -0
- package/src/infra/containers/runtime/development-container +19 -0
- package/src/infra/containers/runtime/noninteractive/index.mjs +7 -0
- package/src/infra/containers/runtime/noninteractive/package.json +6 -0
- package/src/infra/containers/runtime/package.json +10 -0
- package/src/infra/containers/runtime/server.mjs +32 -0
- package/src/infra/containers/transfer.ts +144 -0
- package/src/infra/containers/workspace.ts +304 -0
- package/src/infra/evidence/index.ts +783 -0
- package/src/infra/evidence/tool-calls.ts +49 -0
- package/src/infra/files.ts +74 -0
- package/src/infra/judging/agent.ts +61 -0
- package/src/infra/judging/contract.ts +207 -0
- package/src/infra/judging/evidence-tool.ts +79 -0
- package/src/infra/judging/index.ts +115 -0
- package/src/infra/judging/judge-agent.md +80 -0
- package/src/infra/judging/observer.ts +168 -0
- package/src/infra/judging/submission.ts +138 -0
- package/src/infra/judging/tools.ts +145 -0
- package/src/infra/opencode/archive.ts +81 -0
- package/src/infra/opencode/auth.ts +119 -0
- package/src/infra/opencode/checkpoint.ts +68 -0
- package/src/infra/opencode/host.ts +40 -0
- package/src/infra/opencode/read-recording.ts +42 -0
- package/src/infra/opencode/session.ts +321 -0
- package/src/infra/sqlite/index.ts +636 -0
- package/src/infra/sqlite/schema.ts +57 -0
- package/src/judgment.ts +40 -0
- package/src/runtime.ts +60 -0
- package/src/session-replay.ts +245 -0
- package/src/types.ts +302 -0
- package/src/view.ts +269 -0
- package/viewer/THIRD_PARTY_LICENSES.md +3039 -0
- package/viewer/assets/Inter-c8O0ljhh.ttf +0 -0
- package/viewer/assets/JetBrainsMonoNerdFontMono-Regular-BmWcHh8R.woff2 +0 -0
- package/viewer/assets/KaTeX_AMS-Regular-BQhdFMY1.woff2 +0 -0
- package/viewer/assets/KaTeX_AMS-Regular-DMm9YOAa.woff +0 -0
- package/viewer/assets/KaTeX_AMS-Regular-DRggAlZN.ttf +0 -0
- package/viewer/assets/KaTeX_Caligraphic-Bold-ATXxdsX0.ttf +0 -0
- package/viewer/assets/KaTeX_Caligraphic-Bold-BEiXGLvX.woff +0 -0
- package/viewer/assets/KaTeX_Caligraphic-Bold-Dq_IR9rO.woff2 +0 -0
- package/viewer/assets/KaTeX_Caligraphic-Regular-CTRA-rTL.woff +0 -0
- package/viewer/assets/KaTeX_Caligraphic-Regular-Di6jR-x-.woff2 +0 -0
- package/viewer/assets/KaTeX_Caligraphic-Regular-wX97UBjC.ttf +0 -0
- package/viewer/assets/KaTeX_Fraktur-Bold-BdnERNNW.ttf +0 -0
- package/viewer/assets/KaTeX_Fraktur-Bold-BsDP51OF.woff +0 -0
- package/viewer/assets/KaTeX_Fraktur-Bold-CL6g_b3V.woff2 +0 -0
- package/viewer/assets/KaTeX_Fraktur-Regular-CB_wures.ttf +0 -0
- package/viewer/assets/KaTeX_Fraktur-Regular-CTYiF6lA.woff2 +0 -0
- package/viewer/assets/KaTeX_Fraktur-Regular-Dxdc4cR9.woff +0 -0
- package/viewer/assets/KaTeX_Main-Bold-Cx986IdX.woff2 +0 -0
- package/viewer/assets/KaTeX_Main-Bold-Jm3AIy58.woff +0 -0
- package/viewer/assets/KaTeX_Main-Bold-waoOVXN0.ttf +0 -0
- package/viewer/assets/KaTeX_Main-BoldItalic-DxDJ3AOS.woff2 +0 -0
- package/viewer/assets/KaTeX_Main-BoldItalic-DzxPMmG6.ttf +0 -0
- package/viewer/assets/KaTeX_Main-BoldItalic-SpSLRI95.woff +0 -0
- package/viewer/assets/KaTeX_Main-Italic-3WenGoN9.ttf +0 -0
- package/viewer/assets/KaTeX_Main-Italic-BMLOBm91.woff +0 -0
- package/viewer/assets/KaTeX_Main-Italic-NWA7e6Wa.woff2 +0 -0
- package/viewer/assets/KaTeX_Main-Regular-B22Nviop.woff2 +0 -0
- package/viewer/assets/KaTeX_Main-Regular-Dr94JaBh.woff +0 -0
- package/viewer/assets/KaTeX_Main-Regular-ypZvNtVU.ttf +0 -0
- package/viewer/assets/KaTeX_Math-BoldItalic-B3XSjfu4.ttf +0 -0
- package/viewer/assets/KaTeX_Math-BoldItalic-CZnvNsCZ.woff2 +0 -0
- package/viewer/assets/KaTeX_Math-BoldItalic-iY-2wyZ7.woff +0 -0
- package/viewer/assets/KaTeX_Math-Italic-DA0__PXp.woff +0 -0
- package/viewer/assets/KaTeX_Math-Italic-flOr_0UB.ttf +0 -0
- package/viewer/assets/KaTeX_Math-Italic-t53AETM-.woff2 +0 -0
- package/viewer/assets/KaTeX_SansSerif-Bold-CFMepnvq.ttf +0 -0
- package/viewer/assets/KaTeX_SansSerif-Bold-D1sUS0GD.woff2 +0 -0
- package/viewer/assets/KaTeX_SansSerif-Bold-DbIhKOiC.woff +0 -0
- package/viewer/assets/KaTeX_SansSerif-Italic-C3H0VqGB.woff2 +0 -0
- package/viewer/assets/KaTeX_SansSerif-Italic-DN2j7dab.woff +0 -0
- package/viewer/assets/KaTeX_SansSerif-Italic-YYjJ1zSn.ttf +0 -0
- package/viewer/assets/KaTeX_SansSerif-Regular-BNo7hRIc.ttf +0 -0
- package/viewer/assets/KaTeX_SansSerif-Regular-CS6fqUqJ.woff +0 -0
- package/viewer/assets/KaTeX_SansSerif-Regular-DDBCnlJ7.woff2 +0 -0
- package/viewer/assets/KaTeX_Script-Regular-C5JkGWo-.ttf +0 -0
- package/viewer/assets/KaTeX_Script-Regular-D3wIWfF6.woff2 +0 -0
- package/viewer/assets/KaTeX_Script-Regular-D5yQViql.woff +0 -0
- package/viewer/assets/KaTeX_Size1-Regular-C195tn64.woff +0 -0
- package/viewer/assets/KaTeX_Size1-Regular-Dbsnue_I.ttf +0 -0
- package/viewer/assets/KaTeX_Size1-Regular-mCD8mA8B.woff2 +0 -0
- package/viewer/assets/KaTeX_Size2-Regular-B7gKUWhC.ttf +0 -0
- package/viewer/assets/KaTeX_Size2-Regular-Dy4dx90m.woff2 +0 -0
- package/viewer/assets/KaTeX_Size2-Regular-oD1tc_U0.woff +0 -0
- package/viewer/assets/KaTeX_Size3-Regular-CTq5MqoE.woff +0 -0
- package/viewer/assets/KaTeX_Size3-Regular-DgpXs0kz.ttf +0 -0
- package/viewer/assets/KaTeX_Size4-Regular-BF-4gkZK.woff +0 -0
- package/viewer/assets/KaTeX_Size4-Regular-DWFBv043.ttf +0 -0
- package/viewer/assets/KaTeX_Size4-Regular-Dl5lxZxV.woff2 +0 -0
- package/viewer/assets/KaTeX_Typewriter-Regular-C0xS9mPB.woff +0 -0
- package/viewer/assets/KaTeX_Typewriter-Regular-CO6r4hn1.woff2 +0 -0
- package/viewer/assets/KaTeX_Typewriter-Regular-D3Ib7_Hf.ttf +0 -0
- package/viewer/assets/abap-CLvhMVsD.js +1 -0
- package/viewer/assets/abnfDiagram-VCTEODGH-B1kcYgQG.js +1 -0
- package/viewer/assets/actionscript-3-Cs6JJXT5.js +1 -0
- package/viewer/assets/ada-C5qYipkI.js +1 -0
- package/viewer/assets/ahk-CA9gdOdF.js +1 -0
- package/viewer/assets/ahk2-_6rzW9BS.js +1 -0
- package/viewer/assets/amoled-DB5yy8ad.js +1 -0
- package/viewer/assets/andromeeda-vGVdxbeo.js +1 -0
- package/viewer/assets/angular-html-DrYCUiZv.js +1 -0
- package/viewer/assets/angular-ts-BgjwQn-Z.js +1 -0
- package/viewer/assets/apache-U0d_L8uA.js +1 -0
- package/viewer/assets/apex-KFGFyKwG.js +1 -0
- package/viewer/assets/apl-DlHqe8o4.js +1 -0
- package/viewer/assets/applescript-CCn79oCD.js +1 -0
- package/viewer/assets/ara-4CJ0cIlV.js +1 -0
- package/viewer/assets/arc-C4nf7ZoY.js +1 -0
- package/viewer/assets/architecture-7GRP2DOG-Be250VKZ.js +1 -0
- package/viewer/assets/architectureDiagram-5GKGNRK7-BIWoe0fO.js +36 -0
- package/viewer/assets/array-BifhSqXX.js +1 -0
- package/viewer/assets/asciidoc-SCjQUq34.js +1 -0
- package/viewer/assets/asm-Cmm7eHzH.js +1 -0
- package/viewer/assets/astro-BrlHnIh9.js +1 -0
- package/viewer/assets/aura-BpwPh8WE.js +1 -0
- package/viewer/assets/aurora-x-CDeNXAV0.js +1 -0
- package/viewer/assets/awk-BWXHIvNe.js +1 -0
- package/viewer/assets/ayu-BkUPVVqh.js +1 -0
- package/viewer/assets/ayu-dark-DluEY0Gj.js +1 -0
- package/viewer/assets/ayu-light-C3h-C4tm.js +1 -0
- package/viewer/assets/ayu-mirage-Bqwy1Gya.js +1 -0
- package/viewer/assets/ballerina-B7ZEbQpA.js +1 -0
- package/viewer/assets/bat-qdxAdBWQ.js +1 -0
- package/viewer/assets/beancount-D-usSTwE.js +1 -0
- package/viewer/assets/berry-DKpUyyne.js +1 -0
- package/viewer/assets/bibtex-Ci_nEsc7.js +1 -0
- package/viewer/assets/bicep-CUHmPFLl.js +1 -0
- package/viewer/assets/bird2-CYIUgYqB.js +1 -0
- package/viewer/assets/blade-DOcerOiW.js +1 -0
- package/viewer/assets/blockDiagram-I7D4REHJ-C3x13a9m.js +129 -0
- package/viewer/assets/bsl-_VjlB9mv.js +1 -0
- package/viewer/assets/c-BMbThxEo.js +1 -0
- package/viewer/assets/c3-D0apG41V.js +1 -0
- package/viewer/assets/c4Diagram-7LVT6UL2-BeWktchx.js +38 -0
- package/viewer/assets/cadence-CQ2zXKGN.js +1 -0
- package/viewer/assets/cairo-DLTphjLi.js +1 -0
- package/viewer/assets/carbonfox-Bo2FTw5U.js +1 -0
- package/viewer/assets/catppuccin-DK9M8P0P.js +1 -0
- package/viewer/assets/catppuccin-frappe-C9Q5zVZY.js +1 -0
- package/viewer/assets/catppuccin-frappe-DIKI31zD.js +1 -0
- package/viewer/assets/catppuccin-latte-CtsyrtPq.js +1 -0
- package/viewer/assets/catppuccin-macchiato-DYnJ67qk.js +1 -0
- package/viewer/assets/catppuccin-macchiato-R8nAf1vR.js +1 -0
- package/viewer/assets/catppuccin-mocha-DjuaAJKi.js +1 -0
- package/viewer/assets/channel-DtULXRee.js +1 -0
- package/viewer/assets/chapel-k2cjkSmc.js +1 -0
- package/viewer/assets/chunk-2Q5K7J3B-C1jixKkw.js +1 -0
- package/viewer/assets/chunk-4HAMMTFA-o6YtJsqE.js +62 -0
- package/viewer/assets/chunk-5VM5RSS4-ZNzvKenW.js +15 -0
- package/viewer/assets/chunk-75Z2AOVW-KohQa9Nw.js +2 -0
- package/viewer/assets/chunk-DU6HZSFF-C88wxc6m.js +127 -0
- package/viewer/assets/chunk-F27PBJKO-mO4AP2as.js +1 -0
- package/viewer/assets/chunk-FOHPRMQF-CoA4358r.js +161 -0
- package/viewer/assets/chunk-GMAD6QVW-DzQX2Cld.js +72 -0
- package/viewer/assets/chunk-GVQU2GXP-CS78RfJD.js +1 -0
- package/viewer/assets/chunk-IMKFNOWR-DxS42GUo.js +231 -0
- package/viewer/assets/chunk-JWPE2WC7-DVXcaiue.js +1 -0
- package/viewer/assets/chunk-L3NEJ4N5-DGo-uYLy.js +1 -0
- package/viewer/assets/chunk-OSK3NFVY-BOcXQG16.js +10 -0
- package/viewer/assets/chunk-P2QGCYS3-B7GnZdIQ.js +1 -0
- package/viewer/assets/chunk-POPQ4Y6H-DgBcYHog.js +1 -0
- package/viewer/assets/chunk-PWAF6VOD-DpEq6qHA.js +1 -0
- package/viewer/assets/chunk-SHT3W25Y-CrOGxKM2.js +168 -0
- package/viewer/assets/chunk-SVP7TREG-4HY7Fljj.js +88 -0
- package/viewer/assets/chunk-TICWLB2K-B1Rl31EC.js +206 -0
- package/viewer/assets/chunk-XXDRQBXY-CK8giEW-.js +1 -0
- package/viewer/assets/chunk-Y2CYZVJY-DsF7k-Jl.js +1 -0
- package/viewer/assets/clarity-CtfBBspg.js +1 -0
- package/viewer/assets/classDiagram-ZZMXUADV-Bmqu01Xf.js +1 -0
- package/viewer/assets/classDiagram-v2-VYDZK3BY-Bmqu01Xf.js +1 -0
- package/viewer/assets/clojure-DqKBuwfJ.js +1 -0
- package/viewer/assets/cmake-Bj61d0ZC.js +1 -0
- package/viewer/assets/cobalt2-DQbhSfAX.js +1 -0
- package/viewer/assets/cobol-CupUwvw9.js +1 -0
- package/viewer/assets/codeowners-C8r90Shi.js +1 -0
- package/viewer/assets/codeql-oeQT6MSM.js +1 -0
- package/viewer/assets/coffee-C17pkozF.js +1 -0
- package/viewer/assets/common-lisp-Cv5bFMCO.js +1 -0
- package/viewer/assets/coq-DuBRPxYg.js +1 -0
- package/viewer/assets/cose-bilkent-JH36ORCC-V1OmrmCv.js +1 -0
- package/viewer/assets/cpp-BJwVQVXg.js +1 -0
- package/viewer/assets/crystal-BF-oCN-L.js +1 -0
- package/viewer/assets/csharp-oqKa8noW.js +1 -0
- package/viewer/assets/css-BMgkVI3c.js +1 -0
- package/viewer/assets/csv-Dx-8-gkx.js +1 -0
- package/viewer/assets/cue-CE9AQfxI.js +1 -0
- package/viewer/assets/cursor-t2F-UZH_.js +1 -0
- package/viewer/assets/cynefin-OW5HDTMX-BEjY_gsc.js +1 -0
- package/viewer/assets/cynefinDiagram-5FMLGOSQ-bfSYgr6z.js +62 -0
- package/viewer/assets/cypher-ClKdZ_lG.js +1 -0
- package/viewer/assets/cytoscape.esm-B-1iW2ft.js +321 -0
- package/viewer/assets/d-qD-0Kul2.js +1 -0
- package/viewer/assets/dagre-C-rnJpZF.js +1 -0
- package/viewer/assets/dagre-GXQ25YYZ-VtwTgGfa.js +4 -0
- package/viewer/assets/dark-plus-Cs2F2srj.js +1 -0
- package/viewer/assets/dart-CnvKMtbv.js +1 -0
- package/viewer/assets/dax-BkyTk9wS.js +1 -0
- package/viewer/assets/defaultLocale-BFoDCU3G.js +1 -0
- package/viewer/assets/desktop-Dlh5hvp9.js +1 -0
- package/viewer/assets/diagram-S7CK7UJ4-CHaAAyW7.js +30 -0
- package/viewer/assets/diagram-UQ7AKVKN-CjL65-As.js +41 -0
- package/viewer/assets/diagram-VSXAHHWV-Cnxrk1TJ.js +3 -0
- package/viewer/assets/diagram-VX7I27RA-D78jPXr7.js +24 -0
- package/viewer/assets/diagram-Z3DM3KII-DKjnOaF2.js +24 -0
- package/viewer/assets/diff-woXpYk--.js +1 -0
- package/viewer/assets/dist-BTYPW0dz.js +1 -0
- package/viewer/assets/docker-IyjqRm3v.js +1 -0
- package/viewer/assets/dotenv-_5a1GRtc.js +1 -0
- package/viewer/assets/dracula-BHWKrbxM.js +1 -0
- package/viewer/assets/dracula-C3s659y-.js +1 -0
- package/viewer/assets/dracula-soft-5eyTD99u.js +1 -0
- package/viewer/assets/dream-maker-DW3nJb8Q.js +1 -0
- package/viewer/assets/ebnfDiagram-PWID7BFC-7XiXh5Lx.js +1 -0
- package/viewer/assets/edge-B34N_HjN.js +1 -0
- package/viewer/assets/elixir-CQzWMyKD.js +1 -0
- package/viewer/assets/elm-BNJiG-wg.js +1 -0
- package/viewer/assets/emacs-lisp-c_oH4hRZ.js +1 -0
- package/viewer/assets/erDiagram-RLTQ6QDP-YPFEgZxU.js +99 -0
- package/viewer/assets/erb-D5Rm3UNa.js +1 -0
- package/viewer/assets/erlang-Cphh6RMH.js +1 -0
- package/viewer/assets/eventmodeling-NTZA5JFV-CJCg5f6p.js +1 -0
- package/viewer/assets/everforest-BEfZxV1e.js +1 -0
- package/viewer/assets/everforest-dark-sB-x3p7T.js +1 -0
- package/viewer/assets/everforest-light-Df2xbC6M.js +1 -0
- package/viewer/assets/fennel-DQxkIbk2.js +1 -0
- package/viewer/assets/fish-BJitypiv.js +1 -0
- package/viewer/assets/flexoki-Dj5gUbwm.js +1 -0
- package/viewer/assets/flowDiagram-HODETNUW-BFzq4Zvs.js +1 -0
- package/viewer/assets/fluent-C03EYrpw.js +1 -0
- package/viewer/assets/fortran-fixed-form-DEKoE2YW.js +1 -0
- package/viewer/assets/fortran-free-form-CYNrtFtB.js +1 -0
- package/viewer/assets/fsharp-D13ZGOAj.js +1 -0
- package/viewer/assets/ganttDiagram-EL5Y4UJY-ergSGtTP.js +292 -0
- package/viewer/assets/gdresource-xMC7jr7x.js +1 -0
- package/viewer/assets/gdscript-DYvRD3hg.js +1 -0
- package/viewer/assets/gdshader-CBce3t8t.js +1 -0
- package/viewer/assets/genie-CV2tkWYe.js +1 -0
- package/viewer/assets/gherkin-DExj1W_8.js +1 -0
- package/viewer/assets/git-commit-BSykSTBG.js +1 -0
- package/viewer/assets/git-rebase-BpIMZIRG.js +1 -0
- package/viewer/assets/gitGraph-4MIJSDKK-CZt4REVj.js +1 -0
- package/viewer/assets/gitGraphDiagram-WWUBYQGX-h9QMzC2r.js +106 -0
- package/viewer/assets/github-DukYz3BY.js +1 -0
- package/viewer/assets/github-dark-C-LZuMrd.js +1 -0
- package/viewer/assets/github-dark-default-DXG-b-1a.js +1 -0
- package/viewer/assets/github-dark-dimmed-Bx1FflLF.js +1 -0
- package/viewer/assets/github-dark-high-contrast-B_tTalzw.js +1 -0
- package/viewer/assets/github-light-EUqPIrTm.js +1 -0
- package/viewer/assets/github-light-default-BXViO-2h.js +1 -0
- package/viewer/assets/github-light-high-contrast-B68TUdTA.js +1 -0
- package/viewer/assets/gleam-CSRkHgEL.js +1 -0
- package/viewer/assets/glimmer-js-SeSwns5_.js +1 -0
- package/viewer/assets/glimmer-ts-BExI5tTz.js +1 -0
- package/viewer/assets/glsl-CXikKw2b.js +1 -0
- package/viewer/assets/gn-ilITqXS6.js +1 -0
- package/viewer/assets/gnuplot-7GGW24-e.js +1 -0
- package/viewer/assets/go-rLFTqkRN.js +1 -0
- package/viewer/assets/graphlib-DS17s2tU.js +1 -0
- package/viewer/assets/graphql-6GpQIInt.js +1 -0
- package/viewer/assets/groovy-CacY0gHj.js +1 -0
- package/viewer/assets/gruvbox-dark-hard-C820rvS2.js +1 -0
- package/viewer/assets/gruvbox-dark-medium-BPjhmG05.js +1 -0
- package/viewer/assets/gruvbox-dark-soft-MrdJrrXF.js +1 -0
- package/viewer/assets/gruvbox-light-hard-BC_s9l72.js +1 -0
- package/viewer/assets/gruvbox-light-medium-BAWPOn9u.js +1 -0
- package/viewer/assets/gruvbox-light-soft-BSMLrYjP.js +1 -0
- package/viewer/assets/gruvbox-mhPG2W60.js +1 -0
- package/viewer/assets/hack-i7rizvpB.js +1 -0
- package/viewer/assets/haml-C68gmkTf.js +1 -0
- package/viewer/assets/handlebars-BBn_c0Xt.js +1 -0
- package/viewer/assets/haskell-D8IpX4py.js +1 -0
- package/viewer/assets/haxe-C8oCnwJJ.js +1 -0
- package/viewer/assets/hcl-Dh228itO.js +1 -0
- package/viewer/assets/hjson-CxZEssPk.js +1 -0
- package/viewer/assets/hlsl-Cvrh5tZx.js +1 -0
- package/viewer/assets/horizon-CE9ld1lL.js +1 -0
- package/viewer/assets/horizon-bright-DSNQnXHK.js +1 -0
- package/viewer/assets/houston-CsvMBhTu.js +1 -0
- package/viewer/assets/html-BObmAZQA.js +1 -0
- package/viewer/assets/html-derivative-BjMuR1LM.js +1 -0
- package/viewer/assets/http-BJBk8gt0.js +1 -0
- package/viewer/assets/hurl-D7obCYfg.js +1 -0
- package/viewer/assets/hxml-Dcm88srW.js +1 -0
- package/viewer/assets/hy-CZbG8q4J.js +1 -0
- package/viewer/assets/imba-DsUTQ-LC.js +1 -0
- package/viewer/assets/index-FMtYXcX7.js +793 -0
- package/viewer/assets/index-Wz0qQyIk.css +2 -0
- package/viewer/assets/info-A6RAGUB7-CrzitQIU.js +1 -0
- package/viewer/assets/infoDiagram-27XIBGKW-D2CgRDxu.js +2 -0
- package/viewer/assets/ini-B5eOa1yu.js +1 -0
- package/viewer/assets/init-C-OQMol4.js +1 -0
- package/viewer/assets/ishikawaDiagram-5VMMS53U-BhnQMHOo.js +70 -0
- package/viewer/assets/java-5VJU_EW8.js +1 -0
- package/viewer/assets/javascript-r8UgndxQ.js +1 -0
- package/viewer/assets/jinja-BY2ibHG5.js +1 -0
- package/viewer/assets/jison-C_7YxymZ.js +1 -0
- package/viewer/assets/journeyDiagram-3NMN7TZE-DSBzKQEa.js +139 -0
- package/viewer/assets/json-8e1WlJYD.js +1 -0
- package/viewer/assets/json5-BR5RXkoi.js +1 -0
- package/viewer/assets/jsonc-CYpm1nAK.js +1 -0
- package/viewer/assets/jsonl-CmCQp5Yx.js +1 -0
- package/viewer/assets/jsonnet-CJTPZ8u_.js +1 -0
- package/viewer/assets/jssm-DXw9l8Rf.js +1 -0
- package/viewer/assets/jsx-BYp5GD4Y.js +1 -0
- package/viewer/assets/julia-Bkob1uVX.js +1 -0
- package/viewer/assets/just-CUnSjty-.js +1 -0
- package/viewer/assets/kanagawa-D0YhNhTs.js +1 -0
- package/viewer/assets/kanagawa-dragon-Bozd7klA.js +1 -0
- package/viewer/assets/kanagawa-lotus-BmuOG9ZW.js +1 -0
- package/viewer/assets/kanagawa-wave-BML5mF2J.js +1 -0
- package/viewer/assets/kanban-definition-UXKFOSKX-CotbrQue.js +89 -0
- package/viewer/assets/katex-ZlcWpGUi.js +257 -0
- package/viewer/assets/kdl-CsD5j6eV.js +1 -0
- package/viewer/assets/kotlin-DhhofPvG.js +1 -0
- package/viewer/assets/kusto-C7mF5XQf.js +1 -0
- package/viewer/assets/laserwave-C_8bwKvT.js +1 -0
- package/viewer/assets/latex-UdE3ivqk.js +1 -0
- package/viewer/assets/lean-CewbzKMR.js +1 -0
- package/viewer/assets/less-DVTAwKKz.js +1 -0
- package/viewer/assets/light-plus-DVQuIRkW.js +1 -0
- package/viewer/assets/line-DDh6a6hz.js +1 -0
- package/viewer/assets/linear-BxMsCbKb.js +1 -0
- package/viewer/assets/liquid-WL4MvEIT.js +1 -0
- package/viewer/assets/llvm-DqHOMS-s.js +1 -0
- package/viewer/assets/log-BNLmms1o.js +1 -0
- package/viewer/assets/logo-Cluzi2Zq.js +1 -0
- package/viewer/assets/lua-CtOubZta.js +1 -0
- package/viewer/assets/luau-C2pBCSgt.js +1 -0
- package/viewer/assets/lucent-orng-DLA2Fg5T.js +1 -0
- package/viewer/assets/make-Dixweg8N.js +1 -0
- package/viewer/assets/markdown-BYOwaDjH.js +1 -0
- package/viewer/assets/markdown.worker-DqWDoHOY.js +358 -0
- package/viewer/assets/marko-C0GSQJi3.js +1 -0
- package/viewer/assets/material-BoIUTXek.js +1 -0
- package/viewer/assets/material-theme-Bm3Qr25_.js +1 -0
- package/viewer/assets/material-theme-darker-2IIEA8gg.js +1 -0
- package/viewer/assets/material-theme-lighter-uhdI0v04.js +1 -0
- package/viewer/assets/material-theme-ocean-CHQ94UKr.js +1 -0
- package/viewer/assets/material-theme-palenight-B5W6OYN7.js +1 -0
- package/viewer/assets/matlab-D7qyCx1q.js +1 -0
- package/viewer/assets/matrix-CeUtxxjs.js +1 -0
- package/viewer/assets/mdc-DUivddFc.js +1 -0
- package/viewer/assets/mdx-DQZ5AkYe.js +1 -0
- package/viewer/assets/mercury-DIlqACMb.js +1 -0
- package/viewer/assets/mermaid-BxpfTIzt.js +1 -0
- package/viewer/assets/mermaid-parser.core-FhKE_uv2.js +7 -0
- package/viewer/assets/mermaid.core-BSOWwyvd.js +44 -0
- package/viewer/assets/min-dark-BSWPekZh.js +1 -0
- package/viewer/assets/min-light-DDpmG2fV.js +1 -0
- package/viewer/assets/mindmap-definition-YA3MSWOX-tsLZ_Opn.js +96 -0
- package/viewer/assets/mipsasm-BMqwQI7S.js +1 -0
- package/viewer/assets/mojo-fg40_ntQ.js +1 -0
- package/viewer/assets/monokai-B_bV9PQ9.js +1 -0
- package/viewer/assets/monokai-wFuGqe6w.js +1 -0
- package/viewer/assets/moonbit-DgjMoLWr.js +1 -0
- package/viewer/assets/move-BYnGhZoM.js +1 -0
- package/viewer/assets/narrat-_X_XdTYD.js +1 -0
- package/viewer/assets/nextflow-BJtWHP5T.js +1 -0
- package/viewer/assets/nextflow-groovy-DJMQeKeT.js +1 -0
- package/viewer/assets/nginx-B__INAtC.js +1 -0
- package/viewer/assets/night-owl-DhmEMT88.js +1 -0
- package/viewer/assets/night-owl-light-eJ-hLW7d.js +1 -0
- package/viewer/assets/nightowl-D5RifUq0.js +1 -0
- package/viewer/assets/nim-UEY12GJF.js +1 -0
- package/viewer/assets/nix-IvuFDN5E.js +1 -0
- package/viewer/assets/nord-BUnqvSkw.js +1 -0
- package/viewer/assets/nord-Cb4Vim4T.js +1 -0
- package/viewer/assets/nsis-CjNLYCLg.js +1 -0
- package/viewer/assets/nushell-DHYfO66s.js +1 -0
- package/viewer/assets/objective-c-D1A_Heim.js +1 -0
- package/viewer/assets/objective-cpp-BsSzOQcm.js +1 -0
- package/viewer/assets/ocaml-O90oeIOV.js +1 -0
- package/viewer/assets/odin-B1RWQWA5.js +1 -0
- package/viewer/assets/one-dark-MycZLZtN.js +1 -0
- package/viewer/assets/one-dark-pro-CLwyXe_n.js +1 -0
- package/viewer/assets/one-light-D7Lr4KcI.js +1 -0
- package/viewer/assets/onedarkpro-BWy38d_X.js +1 -0
- package/viewer/assets/opencode-Cy-bxafp.js +1 -0
- package/viewer/assets/openscad-BUDT5pXO.js +1 -0
- package/viewer/assets/ordinal-BDEzSJ7C.js +1 -0
- package/viewer/assets/org-DDIZsz95.js +1 -0
- package/viewer/assets/orng-D2APmIHg.js +1 -0
- package/viewer/assets/osaka-jade-DnfGjKZL.js +1 -0
- package/viewer/assets/packet-AYTQ26CC-CqdJnqfo.js +1 -0
- package/viewer/assets/palenight-DG9raf7e.js +1 -0
- package/viewer/assets/pascal-4ZHwLPI5.js +1 -0
- package/viewer/assets/path-fybaL0A-.js +1 -0
- package/viewer/assets/pegDiagram-XKGWAZYB-BJs7qu7m.js +1 -0
- package/viewer/assets/perl-D5KwNZt7.js +1 -0
- package/viewer/assets/php-BEs7d_Dk.js +1 -0
- package/viewer/assets/pie-WAS4IAKB-lgzqBWqy.js +1 -0
- package/viewer/assets/pieDiagram-E7YTZNPT-Cxe9LnnE.js +39 -0
- package/viewer/assets/pierre-dark-sU4Zdns8.js +1 -0
- package/viewer/assets/pierre-dark-soft-HXgun5vs.js +1 -0
- package/viewer/assets/pierre-dark-vibrant-D4RhcSIK.js +1 -0
- package/viewer/assets/pierre-light-DWBk51d8.js +1 -0
- package/viewer/assets/pierre-light-soft-CodEzPxf.js +1 -0
- package/viewer/assets/pierre-light-vibrant-B76X7i5Y.js +1 -0
- package/viewer/assets/pkl-ot-7Btpt.js +1 -0
- package/viewer/assets/plastic-DQwYfKfQ.js +1 -0
- package/viewer/assets/plsql-DGHpHOYJ.js +1 -0
- package/viewer/assets/po-BiJDBrnU.js +1 -0
- package/viewer/assets/poimandres-DRFjx7u4.js +1 -0
- package/viewer/assets/polar-C7UOKdEL.js +1 -0
- package/viewer/assets/postcss-BXeXVLqQ.js +1 -0
- package/viewer/assets/powerquery-DNMTfnFr.js +1 -0
- package/viewer/assets/powershell-CCSX75yM.js +1 -0
- package/viewer/assets/prisma-DMgNhnYS.js +1 -0
- package/viewer/assets/prolog-iXnhIJG7.js +1 -0
- package/viewer/assets/proto-DB4EqR-F.js +1 -0
- package/viewer/assets/pug-B0dsfQOV.js +1 -0
- package/viewer/assets/puppet-CDv2pdJW.js +1 -0
- package/viewer/assets/purescript-9MfHhQsQ.js +1 -0
- package/viewer/assets/python-gzcpVVnB.js +1 -0
- package/viewer/assets/qml-XPIhrIzf.js +1 -0
- package/viewer/assets/qmldir-DCQb3MpD.js +1 -0
- package/viewer/assets/qss-Fe1Jh2GI.js +1 -0
- package/viewer/assets/quadrantDiagram-AXDQQJYC-D5JGwC11.js +7 -0
- package/viewer/assets/r-C6hQdpIM.js +1 -0
- package/viewer/assets/racket-DcIDlBhZ.js +1 -0
- package/viewer/assets/radar-RG4KPBEZ-B7JRLhUA.js +1 -0
- package/viewer/assets/railroad-74A4TZTK-Df7ZQBLn.js +1 -0
- package/viewer/assets/railroad-abnf-HS5TGJTU-Z8q3QVTf.js +1 -0
- package/viewer/assets/railroad-ebnf-LZEXJU2U-BWoRkG3W.js +1 -0
- package/viewer/assets/railroad-peg-WCYAUIDC-BzoV1cN7.js +1 -0
- package/viewer/assets/railroadDiagram-O6MQD6OU-DZJzpmP8.js +1 -0
- package/viewer/assets/raku-B3gFvitq.js +1 -0
- package/viewer/assets/razor-C3KSUuOw.js +1 -0
- package/viewer/assets/rbs-C0MpOrj4.js +1 -0
- package/viewer/assets/red-DX1BQOs8.js +1 -0
- package/viewer/assets/reg-CRGYupPL.js +1 -0
- package/viewer/assets/regexp-n-T936Rk.js +1 -0
- package/viewer/assets/rel-BtDbiS_P.js +1 -0
- package/viewer/assets/requirementDiagram-BXWQKSXE-Be9lEFrk.js +84 -0
- package/viewer/assets/riscv-Ckw8ddFX.js +1 -0
- package/viewer/assets/ron-VUp2lXgN.js +1 -0
- package/viewer/assets/rose-pine-aMsaMD1D.js +1 -0
- package/viewer/assets/rose-pine-dawn-8PwUyIlP.js +1 -0
- package/viewer/assets/rose-pine-moon-DO-qXg-I.js +1 -0
- package/viewer/assets/rosepine-CVx-z9QY.js +1 -0
- package/viewer/assets/rosmsg-CAekHB0j.js +1 -0
- package/viewer/assets/rough.esm-Dy-Kn_BL.js +1 -0
- package/viewer/assets/rst-fVnHDyXD.js +1 -0
- package/viewer/assets/ruby-DuRgxlXF.js +1 -0
- package/viewer/assets/rust-Cfkwpbl8.js +1 -0
- package/viewer/assets/sankeyDiagram-P5KCCOFB-BsgKAEBN.js +40 -0
- package/viewer/assets/sas-C0WA2OFa.js +1 -0
- package/viewer/assets/sass-DXrisJhu.js +1 -0
- package/viewer/assets/scala-BiFt_7mu.js +1 -0
- package/viewer/assets/scheme-DQCgrYNe.js +1 -0
- package/viewer/assets/scss-CQDZ4smR.js +1 -0
- package/viewer/assets/sdbl-bTVj8UrX.js +1 -0
- package/viewer/assets/sequenceDiagram-WJ2MYXX4-BkUDWL0D.js +162 -0
- package/viewer/assets/shaderlab-TOUzSsQk.js +1 -0
- package/viewer/assets/shadesofpurple-dFPspHnY.js +1 -0
- package/viewer/assets/shellscript-CE84G0GP.js +1 -0
- package/viewer/assets/shellsession-BdJvdaqL.js +1 -0
- package/viewer/assets/sizeCapture-INFHLROL-B0uUizjq.js +1 -0
- package/viewer/assets/slack-dark-DnToyrRv.js +1 -0
- package/viewer/assets/slack-ochin-B2OO5cIa.js +1 -0
- package/viewer/assets/smalltalk-BHpQb5nM.js +1 -0
- package/viewer/assets/smithy-Dgaz7aur.js +1 -0
- package/viewer/assets/snazzy-light-4G7pJPwS.js +1 -0
- package/viewer/assets/solarized-19OwMz-R.js +1 -0
- package/viewer/assets/solarized-dark-UdsmMDCL.js +1 -0
- package/viewer/assets/solarized-light-C1IUL_tW.js +1 -0
- package/viewer/assets/solidity-Bj7E1mRn.js +1 -0
- package/viewer/assets/soy-BSxOREC0.js +1 -0
- package/viewer/assets/sparql-D_iOobhT.js +1 -0
- package/viewer/assets/splunk-BC2Px7Mm.js +1 -0
- package/viewer/assets/sprite-B9ixRF_T.svg +1110 -0
- package/viewer/assets/sprite-Fb-TFjRY.svg +11707 -0
- package/viewer/assets/sql-D4MN9uRI.js +1 -0
- package/viewer/assets/src-BT_kESTA.js +1 -0
- package/viewer/assets/ssh-config-BgfXC-Er.js +1 -0
- package/viewer/assets/stata-B6nXAaGV.js +1 -0
- package/viewer/assets/stateDiagram-D77RDMKH-C80wp8xG.js +1 -0
- package/viewer/assets/stateDiagram-v2-MP3YSRHH-CYdsKKlz.js +1 -0
- package/viewer/assets/stylus-B6D30XZt.js +1 -0
- package/viewer/assets/surrealql-fVjZmhV3.js +1 -0
- package/viewer/assets/svelte-DnBQ5p7d.js +1 -0
- package/viewer/assets/swift-CyEgAFGc.js +1 -0
- package/viewer/assets/swimlanes-42K2YHIH-Cff6yT8e.js +1 -0
- package/viewer/assets/swimlanesDiagram-VR7AAH4N-niTUGjMT.js +8 -0
- package/viewer/assets/synthwave-84-nFMaYfgc.js +1 -0
- package/viewer/assets/synthwave84-D5i2TsvP.js +1 -0
- package/viewer/assets/system-verilog-Dh5ihWbf.js +1 -0
- package/viewer/assets/systemd-BxMlprV5.js +1 -0
- package/viewer/assets/talonscript-CohzipZa.js +1 -0
- package/viewer/assets/tasl-DMoTqEGO.js +1 -0
- package/viewer/assets/tcl-CZd0xW_V.js +1 -0
- package/viewer/assets/templ-TCx51Uts.js +1 -0
- package/viewer/assets/terraform-DswuEJGm.js +1 -0
- package/viewer/assets/tex-CF6SMAQ_.js +1 -0
- package/viewer/assets/timeline-definition-24CTP7MA-Cje7BvaN.js +120 -0
- package/viewer/assets/tokyo-night-oM2G3aXe.js +1 -0
- package/viewer/assets/tokyonight-DLjt3AtF.js +1 -0
- package/viewer/assets/toml-CcmNWLt0.js +1 -0
- package/viewer/assets/treeView-Q6P3EWNA-D-QLY3IO.js +1 -0
- package/viewer/assets/treemap-WGGIJYW6-BCHcTa4N.js +1 -0
- package/viewer/assets/ts-tags-BqLRnADX.js +1 -0
- package/viewer/assets/tsv-sltzmVWM.js +1 -0
- package/viewer/assets/tsx-D7KpDX0b.js +1 -0
- package/viewer/assets/turtle-ByJddavk.js +1 -0
- package/viewer/assets/twig-CEZLObpN.js +1 -0
- package/viewer/assets/typescript-DAbIDLbf.js +1 -0
- package/viewer/assets/typespec-BRdr0IET.js +1 -0
- package/viewer/assets/typst-CqN7-whv.js +1 -0
- package/viewer/assets/v-DBj4K9d_.js +1 -0
- package/viewer/assets/vala-zf12oZj6.js +1 -0
- package/viewer/assets/vb-DOwzaHZ0.js +1 -0
- package/viewer/assets/vennDiagram-4TSXK5OY-Bz20rx02.js +34 -0
- package/viewer/assets/vercel-DiG3Swzk.js +1 -0
- package/viewer/assets/verilog-B2gwz-sJ.js +1 -0
- package/viewer/assets/vesper-BArq_vZH.js +1 -0
- package/viewer/assets/vesper-DdrHHSXu.js +1 -0
- package/viewer/assets/vhdl-BroJfC0k.js +1 -0
- package/viewer/assets/viml-DvXPmvsu.js +1 -0
- package/viewer/assets/vitesse-black-fwtXNY1n.js +1 -0
- package/viewer/assets/vitesse-dark-BZCL-v6S.js +1 -0
- package/viewer/assets/vitesse-light-VbXTXTou.js +1 -0
- package/viewer/assets/vue-html-DQmvvKbd.js +1 -0
- package/viewer/assets/vue-kTxN6wdc.js +1 -0
- package/viewer/assets/vue-vine-Did3qTVW.js +1 -0
- package/viewer/assets/vyper-CgoNMtux.js +1 -0
- package/viewer/assets/wardley-WFR3VGLG-CpXaqOit.js +1 -0
- package/viewer/assets/wardleyDiagram-VM6X3IG4-Cz06jHpA.js +78 -0
- package/viewer/assets/wasm-BnjxR4X6.js +1 -0
- package/viewer/assets/wasm-ByWQv1Qj.js +1 -0
- package/viewer/assets/wenyan-C8pVoKbM.js +1 -0
- package/viewer/assets/wgsl-BsKzXJz4.js +1 -0
- package/viewer/assets/wikitext-ClFFjSW2.js +1 -0
- package/viewer/assets/wit-DdvCle-K.js +1 -0
- package/viewer/assets/wolfram-DLL8P-h_.js +1 -0
- package/viewer/assets/worker-DDDBDFQT.js +160 -0
- package/viewer/assets/xml-te2hFcAA.js +1 -0
- package/viewer/assets/xsl-X4fuKSVl.js +1 -0
- package/viewer/assets/xychartDiagram-S5SC5T6Z-BXCw0sfu.js +7 -0
- package/viewer/assets/yaml-NHtDSZab.js +1 -0
- package/viewer/assets/zenburn-COSjCZYx.js +1 -0
- package/viewer/assets/zenscript-BnlCZFoB.js +1 -0
- package/viewer/assets/zig-CMLA9XwU.js +1 -0
- package/viewer/index.html +14 -0
|
@@ -0,0 +1,218 @@
|
|
|
1
|
+
import { mkdir, mkdtemp, rm } from "node:fs/promises";
|
|
2
|
+
import { tmpdir } from "node:os";
|
|
3
|
+
import { resolve } from "node:path";
|
|
4
|
+
import type {
|
|
5
|
+
BenchmarkDefinition,
|
|
6
|
+
ModelRef,
|
|
7
|
+
PreparationStep,
|
|
8
|
+
SessionArchive,
|
|
9
|
+
EvidenceRef,
|
|
10
|
+
OpenCodeStreamEvent,
|
|
11
|
+
} from "../../types";
|
|
12
|
+
import { CandidateContainer } from "./oci";
|
|
13
|
+
import { EvidenceCapture } from "../evidence";
|
|
14
|
+
import { createSessionDatabase, OPENCODE_VERSION } from "../opencode/host";
|
|
15
|
+
import { credentialsFor, removeCredentials } from "../opencode/auth";
|
|
16
|
+
import { runSession, clientFor, type SessionResult } from "../opencode/session";
|
|
17
|
+
import { readArchivedSession } from "../opencode/archive";
|
|
18
|
+
import { hash, errorMessage, writeJson } from "../files";
|
|
19
|
+
import type { EvidenceFeed } from "../../evidence";
|
|
20
|
+
import { readDurableCheckpoint } from "../opencode/checkpoint";
|
|
21
|
+
|
|
22
|
+
type CandidateInput = {
|
|
23
|
+
model: ModelRef;
|
|
24
|
+
prompt: string;
|
|
25
|
+
workspace: string;
|
|
26
|
+
prepare: readonly PreparationStep[];
|
|
27
|
+
timeoutMs: number;
|
|
28
|
+
websearch: "exa" | false;
|
|
29
|
+
container: BenchmarkDefinition["container"];
|
|
30
|
+
imageId: string;
|
|
31
|
+
providers?: BenchmarkDefinition["candidate"]["providers"];
|
|
32
|
+
};
|
|
33
|
+
type CandidateResult = {
|
|
34
|
+
state: "completed" | "stopped" | "failed" | "timed_out";
|
|
35
|
+
session?: SessionArchive;
|
|
36
|
+
evidence?: EvidenceRef;
|
|
37
|
+
error?: string;
|
|
38
|
+
};
|
|
39
|
+
|
|
40
|
+
/** Owns the container and transport. Judge data cannot enter this interface. */
|
|
41
|
+
export async function executeCandidate(
|
|
42
|
+
input: CandidateInput,
|
|
43
|
+
directory: string,
|
|
44
|
+
onEvent: (event: OpenCodeStreamEvent) => void,
|
|
45
|
+
controls: {
|
|
46
|
+
signal?: AbortSignal;
|
|
47
|
+
onEvidence?: (feed: EvidenceFeed) => void;
|
|
48
|
+
} = {},
|
|
49
|
+
): Promise<CandidateResult> {
|
|
50
|
+
await mkdir(directory, { recursive: true });
|
|
51
|
+
const staging = await mkdtemp(
|
|
52
|
+
resolve(process.env.TMP ?? tmpdir(), "session-"),
|
|
53
|
+
);
|
|
54
|
+
let container: CandidateContainer | undefined,
|
|
55
|
+
captured = false,
|
|
56
|
+
rootId: string | undefined;
|
|
57
|
+
let outcome: SessionResult | undefined;
|
|
58
|
+
let capture: EvidenceCapture | undefined;
|
|
59
|
+
try {
|
|
60
|
+
const seedDatabase = resolve(staging, "opencode.db");
|
|
61
|
+
await createSessionDatabase(
|
|
62
|
+
seedDatabase,
|
|
63
|
+
credentialsFor(input.model, input.websearch),
|
|
64
|
+
);
|
|
65
|
+
container = await CandidateContainer.create(input.container, input.imageId);
|
|
66
|
+
await container.prepare(
|
|
67
|
+
input.workspace,
|
|
68
|
+
seedDatabase,
|
|
69
|
+
input.websearch,
|
|
70
|
+
input.prepare,
|
|
71
|
+
staging,
|
|
72
|
+
input.providers,
|
|
73
|
+
);
|
|
74
|
+
const initial = resolve(staging, "initial");
|
|
75
|
+
await container.snapshot(initial, staging);
|
|
76
|
+
capture = await EvidenceCapture.create(
|
|
77
|
+
resolve(directory, "evidence"),
|
|
78
|
+
initial,
|
|
79
|
+
{ excludeDirectories: ["node_modules"] },
|
|
80
|
+
);
|
|
81
|
+
const recorded = new Set<string>();
|
|
82
|
+
let recordingError: unknown;
|
|
83
|
+
const record = (event: OpenCodeStreamEvent) => {
|
|
84
|
+
if (!("id" in event) || recorded.has(event.id)) return;
|
|
85
|
+
if (event.type === "session.created" && !event.data.parentID)
|
|
86
|
+
rootId ??= event.data.sessionID;
|
|
87
|
+
try {
|
|
88
|
+
capture!.event(event);
|
|
89
|
+
onEvent(event);
|
|
90
|
+
recorded.add(event.id);
|
|
91
|
+
} catch (error) {
|
|
92
|
+
recordingError = error;
|
|
93
|
+
throw error;
|
|
94
|
+
}
|
|
95
|
+
};
|
|
96
|
+
const url = await container.start(input.timeoutMs);
|
|
97
|
+
const client = clientFor(url, container.password);
|
|
98
|
+
let transportError: string | undefined;
|
|
99
|
+
try {
|
|
100
|
+
outcome = await runSession(
|
|
101
|
+
client,
|
|
102
|
+
{
|
|
103
|
+
model: input.model,
|
|
104
|
+
agent: "build",
|
|
105
|
+
prompt: input.prompt,
|
|
106
|
+
directory: "/workspace",
|
|
107
|
+
timeoutMs: input.timeoutMs,
|
|
108
|
+
signal: controls.signal,
|
|
109
|
+
onCreated: (id) => {
|
|
110
|
+
rootId = id;
|
|
111
|
+
controls.onEvidence?.(
|
|
112
|
+
capture!.feed(input.prompt, (cutoff, signal) =>
|
|
113
|
+
readDurableCheckpoint(client, id, cutoff, signal, record),
|
|
114
|
+
),
|
|
115
|
+
);
|
|
116
|
+
},
|
|
117
|
+
},
|
|
118
|
+
record,
|
|
119
|
+
);
|
|
120
|
+
} catch (error) {
|
|
121
|
+
if (recordingError || !rootId) throw error;
|
|
122
|
+
transportError = errorMessage(error);
|
|
123
|
+
// The independent server deadline can close HTTP before final reads.
|
|
124
|
+
// Read the stopped, scrubbed database before deciding the outcome.
|
|
125
|
+
}
|
|
126
|
+
const deadlineExpired =
|
|
127
|
+
container.deadlineAt !== undefined && Date.now() >= container.deadlineAt;
|
|
128
|
+
capture.close();
|
|
129
|
+
await container.stop();
|
|
130
|
+
const database = resolve(directory, "opencode.db");
|
|
131
|
+
await container.saveDatabase(database);
|
|
132
|
+
removeCredentials(database);
|
|
133
|
+
if (!outcome || controls.onEvidence) {
|
|
134
|
+
const archived = await readArchivedSession(database, rootId!, record);
|
|
135
|
+
const previous = outcome;
|
|
136
|
+
outcome = previous
|
|
137
|
+
? { ...archived.result, state: previous.state, error: previous.error }
|
|
138
|
+
: archived.result;
|
|
139
|
+
if (!archived.outcome) outcome.error = transportError ?? outcome.error;
|
|
140
|
+
if (
|
|
141
|
+
!previous &&
|
|
142
|
+
(archived.outcome === "interrupted" || !archived.outcome) &&
|
|
143
|
+
deadlineExpired
|
|
144
|
+
) {
|
|
145
|
+
outcome.state = "timed_out";
|
|
146
|
+
outcome.error = "Candidate session exceeded its time limit";
|
|
147
|
+
} else if (
|
|
148
|
+
!previous &&
|
|
149
|
+
(archived.outcome === "interrupted" || !archived.outcome) &&
|
|
150
|
+
controls.signal?.aborted
|
|
151
|
+
) {
|
|
152
|
+
outcome.state = "stopped";
|
|
153
|
+
outcome.error = undefined;
|
|
154
|
+
}
|
|
155
|
+
}
|
|
156
|
+
const final = resolve(directory, "workspace");
|
|
157
|
+
await container.snapshot(final, staging);
|
|
158
|
+
await writeJson(resolve(directory, "session.json"), outcome.sessions);
|
|
159
|
+
const manifest = await capture.finish({
|
|
160
|
+
prompt: input.prompt,
|
|
161
|
+
response: { text: outcome.text },
|
|
162
|
+
tools: outcome.tools,
|
|
163
|
+
workspace: final,
|
|
164
|
+
});
|
|
165
|
+
const session: SessionArchive = {
|
|
166
|
+
sessionId: outcome.sessions[0]!.info.id,
|
|
167
|
+
database: "opencode.db",
|
|
168
|
+
databaseHash: hash(await Bun.file(database).bytes()),
|
|
169
|
+
opencodeVersion: OPENCODE_VERSION,
|
|
170
|
+
accounting: outcome.accounting,
|
|
171
|
+
};
|
|
172
|
+
captured = true;
|
|
173
|
+
return {
|
|
174
|
+
state: outcome.state,
|
|
175
|
+
session,
|
|
176
|
+
evidence: { directory: "evidence", hash: manifest.sha256 },
|
|
177
|
+
error: outcome.error,
|
|
178
|
+
};
|
|
179
|
+
} catch (error) {
|
|
180
|
+
await container?.stop().catch(() => {});
|
|
181
|
+
let session: SessionArchive | undefined;
|
|
182
|
+
if (container) {
|
|
183
|
+
const database = resolve(directory, "opencode.db");
|
|
184
|
+
try {
|
|
185
|
+
await container.saveDatabase(database);
|
|
186
|
+
removeCredentials(database);
|
|
187
|
+
if (rootId)
|
|
188
|
+
session = {
|
|
189
|
+
sessionId: rootId,
|
|
190
|
+
database: "opencode.db",
|
|
191
|
+
databaseHash: hash(await Bun.file(database).bytes()),
|
|
192
|
+
opencodeVersion: OPENCODE_VERSION,
|
|
193
|
+
accounting: outcome?.accounting,
|
|
194
|
+
};
|
|
195
|
+
} catch (archiveError) {
|
|
196
|
+
await writeJson(resolve(directory, "archive-error.json"), {
|
|
197
|
+
error: errorMessage(archiveError),
|
|
198
|
+
});
|
|
199
|
+
}
|
|
200
|
+
}
|
|
201
|
+
if (outcome)
|
|
202
|
+
await writeJson(resolve(directory, "candidate-checkpoint.json"), outcome);
|
|
203
|
+
return {
|
|
204
|
+
state: "failed",
|
|
205
|
+
session,
|
|
206
|
+
error: `${errorMessage(error)}${container ? ` (workspace retained in ${container.name})` : ""}`,
|
|
207
|
+
};
|
|
208
|
+
} finally {
|
|
209
|
+
capture?.close();
|
|
210
|
+
if (captured) await container?.close();
|
|
211
|
+
await rm(staging, {
|
|
212
|
+
recursive: true,
|
|
213
|
+
force: true,
|
|
214
|
+
maxRetries: 5,
|
|
215
|
+
retryDelay: 100,
|
|
216
|
+
});
|
|
217
|
+
}
|
|
218
|
+
}
|
|
@@ -0,0 +1,31 @@
|
|
|
1
|
+
import type { Engine } from "../../types";
|
|
2
|
+
|
|
3
|
+
/** The engine adapter owns argv, process cancellation, and diagnostics. */
|
|
4
|
+
export async function engineCommand(
|
|
5
|
+
engine: Engine,
|
|
6
|
+
args: string[],
|
|
7
|
+
options: { output?: string; input?: string; timeoutMs?: number } = {},
|
|
8
|
+
) {
|
|
9
|
+
const child = Bun.spawn([engine, ...args], {
|
|
10
|
+
stdout: "pipe",
|
|
11
|
+
stderr: "pipe",
|
|
12
|
+
stdin: options.input ? Bun.file(options.input) : "ignore",
|
|
13
|
+
});
|
|
14
|
+
const timer = setTimeout(() => child.kill(), options.timeoutMs ?? 120_000);
|
|
15
|
+
try {
|
|
16
|
+
const [code, stdout, stderr] = await Promise.all([
|
|
17
|
+
child.exited,
|
|
18
|
+
options.output
|
|
19
|
+
? Bun.write(options.output, new Response(child.stdout)).then(() => "")
|
|
20
|
+
: new Response(child.stdout).text(),
|
|
21
|
+
new Response(child.stderr).text(),
|
|
22
|
+
]);
|
|
23
|
+
if (code !== 0)
|
|
24
|
+
throw new Error(
|
|
25
|
+
`${engine} ${args[0]} failed: ${stderr.trim() || stdout.trim() || code}`,
|
|
26
|
+
);
|
|
27
|
+
return stdout.trim();
|
|
28
|
+
} finally {
|
|
29
|
+
clearTimeout(timer);
|
|
30
|
+
}
|
|
31
|
+
}
|
|
@@ -0,0 +1,335 @@
|
|
|
1
|
+
import { mkdir, rm, stat } from "node:fs/promises";
|
|
2
|
+
import { resolve, posix, basename, dirname } from "node:path";
|
|
3
|
+
import { randomUUID } from "node:crypto";
|
|
4
|
+
import { fileURLToPath } from "node:url";
|
|
5
|
+
import {
|
|
6
|
+
CANDIDATE_TIMEOUT_MS,
|
|
7
|
+
type BenchmarkDefinition,
|
|
8
|
+
type PreparationStep,
|
|
9
|
+
} from "../../types";
|
|
10
|
+
import { engineCommand } from "./docker";
|
|
11
|
+
import { extractWorkspaceArchive } from "./transfer";
|
|
12
|
+
import { treeHash, fingerprint, writeJson } from "../files";
|
|
13
|
+
import { OPENCODE_VERSION } from "../opencode/host";
|
|
14
|
+
import { judgingFingerprint } from "../judging";
|
|
15
|
+
|
|
16
|
+
const runtimeDirectory = fileURLToPath(new URL("./runtime/", import.meta.url));
|
|
17
|
+
export async function buildImage(config: BenchmarkDefinition["container"]) {
|
|
18
|
+
const source = await treeHash(runtimeDirectory);
|
|
19
|
+
await engineCommand(
|
|
20
|
+
config.engine,
|
|
21
|
+
[
|
|
22
|
+
"build",
|
|
23
|
+
"--label",
|
|
24
|
+
`openeval.runtime=${source}`,
|
|
25
|
+
"-t",
|
|
26
|
+
config.image,
|
|
27
|
+
runtimeDirectory,
|
|
28
|
+
],
|
|
29
|
+
{ timeoutMs: 600_000 },
|
|
30
|
+
);
|
|
31
|
+
}
|
|
32
|
+
export async function inspectImage(config: BenchmarkDefinition["container"]) {
|
|
33
|
+
const inspected = await engineCommand(config.engine, [
|
|
34
|
+
"image",
|
|
35
|
+
"inspect",
|
|
36
|
+
config.image,
|
|
37
|
+
"--format",
|
|
38
|
+
'{{.Id}}|{{index .Config.Labels "openeval.runtime"}}',
|
|
39
|
+
]);
|
|
40
|
+
const [id, source] = inspected.split("|");
|
|
41
|
+
if (source !== (await treeHash(runtimeDirectory)))
|
|
42
|
+
throw new Error(
|
|
43
|
+
"Candidate runtime changed or the image is not initialized. Run the image command first.",
|
|
44
|
+
);
|
|
45
|
+
return id;
|
|
46
|
+
}
|
|
47
|
+
export async function runtimeFingerprint(
|
|
48
|
+
config: BenchmarkDefinition["container"],
|
|
49
|
+
) {
|
|
50
|
+
const imageId = await inspectImage(config);
|
|
51
|
+
return {
|
|
52
|
+
imageId,
|
|
53
|
+
candidateHash: fingerprint({
|
|
54
|
+
imageId,
|
|
55
|
+
sdk: OPENCODE_VERSION,
|
|
56
|
+
engine: config.engine,
|
|
57
|
+
cpus: config.cpus,
|
|
58
|
+
memoryMiB: config.memoryMiB,
|
|
59
|
+
implementation: await treeHash(resolve(runtimeDirectory, ".."), {
|
|
60
|
+
ignore: /\.test\.[jt]sx?$/,
|
|
61
|
+
}),
|
|
62
|
+
sessions: await treeHash(
|
|
63
|
+
fileURLToPath(new URL("../opencode/", import.meta.url)),
|
|
64
|
+
{ ignore: /\.test\.[jt]sx?$/ },
|
|
65
|
+
),
|
|
66
|
+
evidence: await treeHash(
|
|
67
|
+
fileURLToPath(new URL("../evidence/", import.meta.url)),
|
|
68
|
+
),
|
|
69
|
+
useCases: await Promise.all(
|
|
70
|
+
[
|
|
71
|
+
"run-eval.ts",
|
|
72
|
+
"run-eval-pipeline.ts",
|
|
73
|
+
"watch-evidence.ts",
|
|
74
|
+
"execution-budget.ts",
|
|
75
|
+
].map((file) =>
|
|
76
|
+
Bun.file(
|
|
77
|
+
fileURLToPath(new URL(`../../app/${file}`, import.meta.url)),
|
|
78
|
+
).text(),
|
|
79
|
+
),
|
|
80
|
+
),
|
|
81
|
+
}),
|
|
82
|
+
judgeHash: await judgingFingerprint(),
|
|
83
|
+
};
|
|
84
|
+
}
|
|
85
|
+
|
|
86
|
+
/** A private workspace and a normal OpenCode installation; no evaluator inputs or code. */
|
|
87
|
+
export class CandidateContainer {
|
|
88
|
+
readonly name = `workspace-${randomUUID()}`;
|
|
89
|
+
readonly password = randomUUID();
|
|
90
|
+
deadlineAt?: number;
|
|
91
|
+
private workspaceVolume = `${this.name}-files`;
|
|
92
|
+
private homeVolume = `${this.name}-home`;
|
|
93
|
+
private stopped = false;
|
|
94
|
+
private removed = false;
|
|
95
|
+
private constructor(
|
|
96
|
+
private config: BenchmarkDefinition["container"],
|
|
97
|
+
readonly imageId: string,
|
|
98
|
+
) {}
|
|
99
|
+
private command(
|
|
100
|
+
args: string[],
|
|
101
|
+
options?: Parameters<typeof engineCommand>[2],
|
|
102
|
+
) {
|
|
103
|
+
return engineCommand(this.config.engine, args, options);
|
|
104
|
+
}
|
|
105
|
+
|
|
106
|
+
static async create(
|
|
107
|
+
config: BenchmarkDefinition["container"],
|
|
108
|
+
imageId: string,
|
|
109
|
+
) {
|
|
110
|
+
const container = new CandidateContainer(config, imageId);
|
|
111
|
+
try {
|
|
112
|
+
for (const volume of [container.workspaceVolume, container.homeVolume])
|
|
113
|
+
await container.command(["volume", "create", volume]);
|
|
114
|
+
await container.command([
|
|
115
|
+
"run",
|
|
116
|
+
"--detach",
|
|
117
|
+
"--init",
|
|
118
|
+
"--name",
|
|
119
|
+
container.name,
|
|
120
|
+
"--hostname",
|
|
121
|
+
"workstation",
|
|
122
|
+
"--cap-drop=ALL",
|
|
123
|
+
"--cap-add=NET_ADMIN",
|
|
124
|
+
"--cap-add=SETUID",
|
|
125
|
+
"--cap-add=SETGID",
|
|
126
|
+
"--cap-add=SETPCAP",
|
|
127
|
+
"--security-opt=no-new-privileges",
|
|
128
|
+
"--pids-limit",
|
|
129
|
+
"256",
|
|
130
|
+
"--cpus",
|
|
131
|
+
String(config.cpus),
|
|
132
|
+
"--memory",
|
|
133
|
+
`${config.memoryMiB}m`,
|
|
134
|
+
"--publish",
|
|
135
|
+
"127.0.0.1::4096",
|
|
136
|
+
"--mount",
|
|
137
|
+
`type=volume,source=${container.workspaceVolume},target=/workspace`,
|
|
138
|
+
"--mount",
|
|
139
|
+
`type=volume,source=${container.homeVolume},target=/home/dev`,
|
|
140
|
+
imageId,
|
|
141
|
+
]);
|
|
142
|
+
return container;
|
|
143
|
+
} catch (error) {
|
|
144
|
+
await container.close();
|
|
145
|
+
throw error;
|
|
146
|
+
}
|
|
147
|
+
}
|
|
148
|
+
async prepare(
|
|
149
|
+
workspace: string,
|
|
150
|
+
database: string,
|
|
151
|
+
websearch: "exa" | false,
|
|
152
|
+
steps: readonly PreparationStep[],
|
|
153
|
+
staging: string,
|
|
154
|
+
providers?: BenchmarkDefinition["candidate"]["providers"],
|
|
155
|
+
) {
|
|
156
|
+
await this.upload(workspace, "/workspace", staging);
|
|
157
|
+
await this.upload(database, "/home/dev/.local/share/opencode", staging);
|
|
158
|
+
const config = resolve(staging, "opencode.json");
|
|
159
|
+
await writeJson(config, {
|
|
160
|
+
plugins: ["/opt/opencode/noninteractive"],
|
|
161
|
+
permissions: [{ action: "*", resource: "*", effect: "allow" }],
|
|
162
|
+
websearch: websearch ? { provider: websearch } : false,
|
|
163
|
+
...(providers ? { providers } : {}),
|
|
164
|
+
});
|
|
165
|
+
await this.upload(config, "/home/dev/.config/opencode", staging);
|
|
166
|
+
for (const step of steps) {
|
|
167
|
+
const workdir = posix.resolve(
|
|
168
|
+
"/workspace",
|
|
169
|
+
step.cwd.replaceAll("\\", "/"),
|
|
170
|
+
);
|
|
171
|
+
if (workdir !== "/workspace" && !workdir.startsWith("/workspace/"))
|
|
172
|
+
throw new Error("Preparation directory escapes the workspace");
|
|
173
|
+
await this.command(
|
|
174
|
+
[
|
|
175
|
+
"exec",
|
|
176
|
+
"--user",
|
|
177
|
+
"dev",
|
|
178
|
+
"--workdir",
|
|
179
|
+
workdir,
|
|
180
|
+
this.name,
|
|
181
|
+
...step.argv,
|
|
182
|
+
],
|
|
183
|
+
{ timeoutMs: 180_000 },
|
|
184
|
+
);
|
|
185
|
+
}
|
|
186
|
+
}
|
|
187
|
+
private async upload(source: string, destination: string, staging: string) {
|
|
188
|
+
const directory = (await stat(source)).isDirectory();
|
|
189
|
+
const archive = resolve(staging, `transfer-${randomUUID()}.tar`);
|
|
190
|
+
try {
|
|
191
|
+
const child = Bun.spawn(
|
|
192
|
+
[
|
|
193
|
+
"tar",
|
|
194
|
+
"-cf",
|
|
195
|
+
archive,
|
|
196
|
+
"-C",
|
|
197
|
+
directory ? source : dirname(source),
|
|
198
|
+
directory ? "." : basename(source),
|
|
199
|
+
],
|
|
200
|
+
{ stdout: "ignore", stderr: "pipe" },
|
|
201
|
+
);
|
|
202
|
+
const [code, error] = await Promise.all([
|
|
203
|
+
child.exited,
|
|
204
|
+
new Response(child.stderr).text(),
|
|
205
|
+
]);
|
|
206
|
+
if (code !== 0) throw new Error(`Workspace transfer failed: ${error}`);
|
|
207
|
+
await this.command(
|
|
208
|
+
[
|
|
209
|
+
"exec",
|
|
210
|
+
"--user",
|
|
211
|
+
"dev",
|
|
212
|
+
"--interactive",
|
|
213
|
+
this.name,
|
|
214
|
+
"tar",
|
|
215
|
+
"--no-same-owner",
|
|
216
|
+
"-xf",
|
|
217
|
+
"-",
|
|
218
|
+
"-C",
|
|
219
|
+
destination,
|
|
220
|
+
],
|
|
221
|
+
{ input: archive },
|
|
222
|
+
);
|
|
223
|
+
} finally {
|
|
224
|
+
await rm(archive, { force: true });
|
|
225
|
+
}
|
|
226
|
+
}
|
|
227
|
+
async start(timeoutMs = CANDIDATE_TIMEOUT_MS) {
|
|
228
|
+
// The server deadline survives a host runner crash.
|
|
229
|
+
this.deadlineAt = Date.now() + Math.ceil(timeoutMs / 1000) * 1000;
|
|
230
|
+
await this.command([
|
|
231
|
+
"exec",
|
|
232
|
+
"--user",
|
|
233
|
+
"dev",
|
|
234
|
+
"--detach",
|
|
235
|
+
"--env",
|
|
236
|
+
`OPENCODE_SERVER_PASSWORD=${this.password}`,
|
|
237
|
+
this.name,
|
|
238
|
+
"sh",
|
|
239
|
+
"-c",
|
|
240
|
+
`exec timeout --signal=TERM --kill-after=5s ${Math.ceil(timeoutMs / 1000)}s bun /opt/opencode/server.mjs > /home/dev/.local/share/opencode/server.log 2>&1`,
|
|
241
|
+
]);
|
|
242
|
+
const port = await this.command(["port", this.name, "4096/tcp"]);
|
|
243
|
+
const url = `http://${port.split(/\r?\n/)[0]}`;
|
|
244
|
+
const deadline = Date.now() + 30_000;
|
|
245
|
+
while (Date.now() < deadline) {
|
|
246
|
+
try {
|
|
247
|
+
if (
|
|
248
|
+
(
|
|
249
|
+
await fetch(`${url}/api/health`, {
|
|
250
|
+
headers: {
|
|
251
|
+
authorization: `Basic ${Buffer.from(`opencode:${this.password}`).toString("base64")}`,
|
|
252
|
+
},
|
|
253
|
+
signal: AbortSignal.timeout(1000),
|
|
254
|
+
})
|
|
255
|
+
).ok
|
|
256
|
+
)
|
|
257
|
+
return url;
|
|
258
|
+
} catch {}
|
|
259
|
+
await Bun.sleep(100);
|
|
260
|
+
}
|
|
261
|
+
const log = await this.inspect([
|
|
262
|
+
"tail",
|
|
263
|
+
"-n",
|
|
264
|
+
"40",
|
|
265
|
+
"/home/dev/.local/share/opencode/server.log",
|
|
266
|
+
]);
|
|
267
|
+
throw new Error(`OpenCode server did not become ready: ${log}`);
|
|
268
|
+
}
|
|
269
|
+
async snapshot(destination: string, staging: string) {
|
|
270
|
+
await mkdir(destination, { recursive: true });
|
|
271
|
+
const archive = resolve(staging, `files-${randomUUID()}.tar`);
|
|
272
|
+
try {
|
|
273
|
+
await this.command(
|
|
274
|
+
[
|
|
275
|
+
"run",
|
|
276
|
+
"--rm",
|
|
277
|
+
"--user",
|
|
278
|
+
"dev",
|
|
279
|
+
"--entrypoint",
|
|
280
|
+
"tar",
|
|
281
|
+
"--mount",
|
|
282
|
+
`type=volume,source=${this.workspaceVolume},target=/source,readonly`,
|
|
283
|
+
this.imageId,
|
|
284
|
+
"--exclude=node_modules",
|
|
285
|
+
"-cf",
|
|
286
|
+
"-",
|
|
287
|
+
"-C",
|
|
288
|
+
"/source",
|
|
289
|
+
".",
|
|
290
|
+
],
|
|
291
|
+
{ output: archive },
|
|
292
|
+
);
|
|
293
|
+
await extractWorkspaceArchive(archive, destination, {
|
|
294
|
+
sourceRoot: "/workspace",
|
|
295
|
+
});
|
|
296
|
+
} finally {
|
|
297
|
+
await rm(archive, { force: true });
|
|
298
|
+
}
|
|
299
|
+
}
|
|
300
|
+
async stop() {
|
|
301
|
+
if (this.stopped) return;
|
|
302
|
+
await this.command(["stop", "--time", "15", this.name], {
|
|
303
|
+
timeoutMs: 30_000,
|
|
304
|
+
});
|
|
305
|
+
this.stopped = true;
|
|
306
|
+
}
|
|
307
|
+
async saveDatabase(destination: string) {
|
|
308
|
+
if (!this.stopped)
|
|
309
|
+
throw new Error("Stop the candidate before archiving its database");
|
|
310
|
+
for (const suffix of ["", "-wal", "-shm"]) {
|
|
311
|
+
try {
|
|
312
|
+
await this.command([
|
|
313
|
+
"cp",
|
|
314
|
+
`${this.name}:/home/dev/.local/share/opencode/opencode.db${suffix}`,
|
|
315
|
+
destination + suffix,
|
|
316
|
+
]);
|
|
317
|
+
} catch (error) {
|
|
318
|
+
if (!suffix) throw error;
|
|
319
|
+
}
|
|
320
|
+
}
|
|
321
|
+
}
|
|
322
|
+
async inspect(command: string[]) {
|
|
323
|
+
return this.command(["exec", "--user", "dev", this.name, ...command]);
|
|
324
|
+
}
|
|
325
|
+
async close() {
|
|
326
|
+
if (this.removed) return;
|
|
327
|
+
this.removed = true;
|
|
328
|
+
await this.command(["rm", "--force", this.name]).catch(() => {});
|
|
329
|
+
for (const volume of [this.workspaceVolume, this.homeVolume])
|
|
330
|
+
await this.command(["volume", "rm", volume]).catch(() => {});
|
|
331
|
+
}
|
|
332
|
+
async [Symbol.asyncDispose]() {
|
|
333
|
+
await this.close();
|
|
334
|
+
}
|
|
335
|
+
}
|
|
@@ -0,0 +1,19 @@
|
|
|
1
|
+
FROM oven/bun:1.4.2
|
|
2
|
+
RUN apt-get update && apt-get install -y --no-install-recommends ca-certificates git ripgrep curl unzip g++ make python3 nodejs npm tar procps \
|
|
3
|
+
&& rm -rf /var/lib/apt/lists/*
|
|
4
|
+
WORKDIR /opt/opencode
|
|
5
|
+
RUN apt-get update && apt-get install -y --no-install-recommends iptables util-linux && rm -rf /var/lib/apt/lists/*
|
|
6
|
+
COPY package.json ./
|
|
7
|
+
RUN bun install --production
|
|
8
|
+
COPY server.mjs ./
|
|
9
|
+
COPY noninteractive ./noninteractive
|
|
10
|
+
COPY development-container /usr/local/bin/development-container
|
|
11
|
+
RUN useradd --create-home --uid 10001 dev \
|
|
12
|
+
&& mkdir -p /workspace /home/dev/.config/opencode /home/dev/.local/share/opencode \
|
|
13
|
+
&& chown -R dev:dev /workspace /home/dev \
|
|
14
|
+
&& git config --system --add safe.directory '*' \
|
|
15
|
+
&& chmod 755 /usr/local/bin/development-container
|
|
16
|
+
ENV HOME=/home/dev
|
|
17
|
+
WORKDIR /workspace
|
|
18
|
+
ENTRYPOINT ["/usr/local/bin/development-container"]
|
|
19
|
+
CMD ["sleep", "infinity"]
|
|
@@ -0,0 +1,19 @@
|
|
|
1
|
+
#!/bin/sh
|
|
2
|
+
set -eu
|
|
3
|
+
|
|
4
|
+
# Allow replies to the local control connection and normal DNS/public egress.
|
|
5
|
+
# A development session has no access to host/LAN services or metadata endpoints.
|
|
6
|
+
iptables -A OUTPUT -m conntrack --ctstate ESTABLISHED,RELATED -j ACCEPT
|
|
7
|
+
iptables -A OUTPUT -p udp --dport 53 -j ACCEPT
|
|
8
|
+
iptables -A OUTPUT -p tcp --dport 53 -j ACCEPT
|
|
9
|
+
for network in 10.0.0.0/8 172.16.0.0/12 192.168.0.0/16 169.254.0.0/16 100.64.0.0/10; do
|
|
10
|
+
iptables -A OUTPUT -d "$network" -j REJECT
|
|
11
|
+
done
|
|
12
|
+
ip6tables -A OUTPUT -m conntrack --ctstate ESTABLISHED,RELATED -j ACCEPT
|
|
13
|
+
ip6tables -A OUTPUT -p udp --dport 53 -j ACCEPT
|
|
14
|
+
ip6tables -A OUTPUT -p tcp --dport 53 -j ACCEPT
|
|
15
|
+
for network in fc00::/7 fe80::/10; do
|
|
16
|
+
ip6tables -A OUTPUT -d "$network" -j REJECT
|
|
17
|
+
done
|
|
18
|
+
|
|
19
|
+
exec setpriv --reuid=dev --regid=dev --init-groups --bounding-set=-all --inh-caps=-all --ambient-caps=-all "$@"
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
import { ServerProcess } from "@opencode/server/process";
|
|
2
|
+
import { Effect, Fiber } from "effect";
|
|
3
|
+
|
|
4
|
+
process.title = "opencode";
|
|
5
|
+
const runtime = Effect.runFork(
|
|
6
|
+
Effect.scoped(
|
|
7
|
+
Effect.gen(function* () {
|
|
8
|
+
yield* ServerProcess.start({
|
|
9
|
+
hostname: "0.0.0.0",
|
|
10
|
+
port: 4096,
|
|
11
|
+
password: process.env.OPENCODE_SERVER_PASSWORD,
|
|
12
|
+
app: { name: "opencode", version: "0.0.0-beta-19296" },
|
|
13
|
+
database: { path: "/home/dev/.local/share/opencode/opencode.db" },
|
|
14
|
+
events: { persist: true },
|
|
15
|
+
});
|
|
16
|
+
yield* Effect.never;
|
|
17
|
+
}),
|
|
18
|
+
),
|
|
19
|
+
);
|
|
20
|
+
let closing = false;
|
|
21
|
+
for (const signal of ["SIGTERM", "SIGINT"])
|
|
22
|
+
process.on(signal, () => {
|
|
23
|
+
if (closing) return;
|
|
24
|
+
closing = true;
|
|
25
|
+
void Effect.runPromise(Fiber.interrupt(runtime)).finally(() =>
|
|
26
|
+
process.exit(0),
|
|
27
|
+
);
|
|
28
|
+
});
|
|
29
|
+
void Effect.runPromise(Fiber.join(runtime)).catch((error) => {
|
|
30
|
+
console.error(error);
|
|
31
|
+
process.exit(1);
|
|
32
|
+
});
|