feihong-code 0.2.2 → 0.5.1
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/.github/FUNDING.yml +4 -0
- package/.github/ISSUE_TEMPLATE/bug_report.md +37 -37
- package/.github/ISSUE_TEMPLATE/config.yml +14 -14
- package/.github/ISSUE_TEMPLATE/feature_request.md +28 -28
- package/.github/PULL_REQUEST_TEMPLATE.md +46 -46
- package/.github/SECURITY.md +67 -67
- package/.github/dependabot.yml +16 -0
- package/.github/workflows/ci.yml +1 -1
- package/AGENT-GUIDE.md +382 -237
- package/CHANGELOG.md +405 -73
- package/CODE_OF_CONDUCT.md +56 -56
- package/CONTRIBUTING.md +68 -68
- package/LICENSE +22 -22
- package/README.md +31 -25
- package/README.self-evolve.md +274 -0
- package/dist/agent/code-review.js +2 -0
- package/dist/agent/code-review.js.map +1 -1
- package/dist/agent/code-writer.js +70 -17
- package/dist/agent/code-writer.js.map +1 -1
- package/dist/agent/context-compactor.js +13 -4
- package/dist/agent/context-compactor.js.map +1 -1
- package/dist/agent/experience.js +262 -62
- package/dist/agent/experience.js.map +1 -1
- package/dist/agent/orchestrator.js +148 -68
- package/dist/agent/orchestrator.js.map +1 -1
- package/dist/agent/quality-gate.js +10 -4
- package/dist/agent/quality-gate.js.map +1 -1
- package/dist/agent/repo-context.js +181 -0
- package/dist/agent/repo-context.js.map +1 -0
- package/dist/agent/repo-reader.js +92 -99
- package/dist/agent/repo-reader.js.map +1 -1
- package/dist/agent/self-heal.js +38 -48
- package/dist/agent/self-heal.js.map +1 -1
- package/dist/agent/self-improver.js +65 -61
- package/dist/agent/self-improver.js.map +1 -1
- package/dist/agent/subagent-summary.js +31 -0
- package/dist/agent/subagent-summary.js.map +1 -0
- package/dist/agent/subagent.js +52 -2
- package/dist/agent/subagent.js.map +1 -1
- package/dist/agent/symbol-index.js +160 -0
- package/dist/agent/symbol-index.js.map +1 -0
- package/dist/agent/team.js +193 -0
- package/dist/agent/team.js.map +1 -0
- package/dist/cli/commands.js +124 -143
- package/dist/cli/commands.js.map +1 -1
- package/dist/cli/index.js +113 -106
- package/dist/cli/index.js.map +1 -1
- package/dist/cli/repl.js +82 -7
- package/dist/cli/repl.js.map +1 -1
- package/dist/cli/run.js +601 -95
- package/dist/cli/run.js.map +1 -1
- package/dist/cli/tui.js +145 -0
- package/dist/cli/tui.js.map +1 -0
- package/dist/cli/version.js +1 -1
- package/dist/enterprise/audit.js +75 -12
- package/dist/enterprise/audit.js.map +1 -1
- package/dist/enterprise/index.js +14 -11
- package/dist/enterprise/index.js.map +1 -1
- package/dist/enterprise/policy.js +22 -10
- package/dist/enterprise/policy.js.map +1 -1
- package/dist/harness/executor.js +127 -0
- package/dist/harness/executor.js.map +1 -0
- package/dist/harness/harness.js +87 -0
- package/dist/harness/harness.js.map +1 -0
- package/dist/harness/index.js +31 -0
- package/dist/harness/index.js.map +1 -0
- package/dist/harness/loader.js +138 -0
- package/dist/harness/loader.js.map +1 -0
- package/dist/harness/reporter.js +34 -0
- package/dist/harness/reporter.js.map +1 -0
- package/dist/harness/types.js +10 -0
- package/dist/harness/types.js.map +1 -0
- package/dist/harness/verifier.js +48 -0
- package/dist/harness/verifier.js.map +1 -0
- package/dist/hello.js +14 -0
- package/dist/hello.js.map +1 -0
- package/dist/memory/auto-summarize.js +208 -0
- package/dist/memory/auto-summarize.js.map +1 -0
- package/dist/memory/index.js +228 -0
- package/dist/memory/index.js.map +1 -0
- package/dist/models/model-router.js +23 -10
- package/dist/models/model-router.js.map +1 -1
- package/dist/plugins/plugin-loader.js +179 -0
- package/dist/plugins/plugin-loader.js.map +1 -0
- package/dist/runtime/event-log.js.map +1 -1
- package/dist/runtime/hooks.js +80 -0
- package/dist/runtime/hooks.js.map +1 -0
- package/dist/self-evolve/hook.js +60 -0
- package/dist/self-evolve/hook.js.map +1 -0
- package/dist/shared/config.js +35 -3
- package/dist/shared/config.js.map +1 -1
- package/dist/shared/i18n.js +535 -0
- package/dist/shared/i18n.js.map +1 -0
- package/dist/skills/grill.js +2 -1
- package/dist/skills/grill.js.map +1 -1
- package/dist/skills/self-heal.js +73 -0
- package/dist/skills/self-heal.js.map +1 -0
- package/dist/skills/skill-loader.js +131 -0
- package/dist/skills/skill-loader.js.map +1 -0
- package/dist/skills/skill-market.js +195 -0
- package/dist/skills/skill-market.js.map +1 -0
- package/dist/tools/analysis/code-analyzer.js +45 -18
- package/dist/tools/analysis/code-analyzer.js.map +1 -1
- package/dist/tools/index.js +5 -0
- package/dist/tools/index.js.map +1 -1
- package/dist/tools/mcp/index.js +84 -0
- package/dist/tools/mcp/index.js.map +1 -0
- package/dist/tools/mcp/mcp-client.js +172 -0
- package/dist/tools/mcp/mcp-client.js.map +1 -0
- package/dist/tools/sandbox.js +126 -0
- package/dist/tools/sandbox.js.map +1 -0
- package/dist/tools/shell/exec.js +25 -0
- package/dist/tools/shell/exec.js.map +1 -1
- package/dist/tools/shell/run-shell.tool.js +4 -1
- package/dist/tools/shell/run-shell.tool.js.map +1 -1
- package/dist/tools/skills/load-skill.tool.js +36 -0
- package/dist/tools/skills/load-skill.tool.js.map +1 -0
- package/dist/tools/tool.interface.js.map +1 -1
- package/dist/tools/tool.registry.js +37 -1
- package/dist/tools/tool.registry.js.map +1 -1
- package/dist/tools/web/web.tool.js +139 -0
- package/dist/tools/web/web.tool.js.map +1 -0
- package/dist/web/auth.js +42 -3
- package/dist/web/auth.js.map +1 -1
- package/dist/web/channels.js +171 -0
- package/dist/web/channels.js.map +1 -0
- package/dist/web/public/index.html +3253 -39
- package/dist/web/public/index.html.tmp +3117 -0
- package/dist/web/public/index_new.html +3196 -0
- package/dist/web/server.js +790 -11
- package/dist/web/server.js.map +1 -1
- package/dist/web/task-queue.js +352 -0
- package/dist/web/task-queue.js.map +1 -0
- package/dist/web/web-config.js +143 -0
- package/dist/web/web-config.js.map +1 -0
- package/docs/Deployment_Guide_EN.md +288 -0
- package/docs/Technical_Manual_EN.md +216 -0
- package/docs/User_Manual_EN.md +314 -0
- package/docs/self-evolve-implementation.md +165 -0
- package/docs/self-evolve.md +166 -0
- package/docs//344/272/247/345/223/201/345/274/200/345/217/221/346/226/207/346/241/243.md +614 -614
- package/docs//344/274/201/344/270/232/351/203/250/347/275/262/344/270/216/345/220/210/350/247/204.md +267 -267
- package/docs//344/275/277/347/224/250/350/257/264/346/230/216/344/271/246.md +318 -358
- package/docs//345/270/270/350/247/201/351/227/256/351/242/230/344/270/216/346/225/205/351/232/234/346/216/222/346/237/245.md +130 -130
- package/docs//346/212/200/346/234/257/350/257/264/346/230/216/344/271/246.md +216 -303
- package/docs//346/236/266/346/236/204/344/270/216API.md +238 -238
- package/docs//347/224/250/346/210/267/346/211/213/345/206/214.md +196 -196
- package/docs//351/203/250/347/275/262/346/214/207/345/215/227.md +165 -165
- package/docs//351/203/250/347/275/262/350/257/264/346/230/216/344/271/246.md +288 -0
- package/docs//351/205/215/347/275/256/345/217/202/350/200/203.md +127 -127
- package/package.json +114 -109
- package/tool-schema.json +127 -127
|
@@ -24,6 +24,7 @@ const fs_1 = require("fs");
|
|
|
24
24
|
const path_1 = require("path");
|
|
25
25
|
const logger_1 = require("../shared/logger");
|
|
26
26
|
const tenant_1 = require("./tenant");
|
|
27
|
+
const i18n_1 = require("../shared/i18n");
|
|
27
28
|
exports.DEFAULT_POLICY = {
|
|
28
29
|
version: 1,
|
|
29
30
|
roles: {
|
|
@@ -225,17 +226,28 @@ function evaluate(policy, input) {
|
|
|
225
226
|
function renderPolicy(policy, role) {
|
|
226
227
|
const rp = policy.roles[role];
|
|
227
228
|
const lines = [
|
|
228
|
-
|
|
229
|
-
|
|
230
|
-
|
|
231
|
-
|
|
232
|
-
|
|
233
|
-
|
|
234
|
-
|
|
235
|
-
|
|
229
|
+
(0, i18n_1.t)('policy.version', { v: policy.version }),
|
|
230
|
+
(0, i18n_1.t)('policy.role', { v: role }),
|
|
231
|
+
(0, i18n_1.t)('policy.allow', { v: rp.allowTools.join(', ') }),
|
|
232
|
+
(0, i18n_1.t)('policy.approval', { v: rp.approvalTools.join(', ') || (0, i18n_1.t)('enterprise.none') }),
|
|
233
|
+
(0, i18n_1.t)('policy.taskCap', { v: rp.maxCostUsd > 0 ? `$${rp.maxCostUsd}` : (0, i18n_1.t)('enterprise.unlimited') }),
|
|
234
|
+
(0, i18n_1.t)('policy.tenantCap', { v: policy.tenantDailyBudgetUsd > 0 ? `$${policy.tenantDailyBudgetUsd}` : (0, i18n_1.t)('enterprise.unlimited') }),
|
|
235
|
+
(0, i18n_1.t)('policy.denyShell', {
|
|
236
|
+
n: policy.denyShell.length,
|
|
237
|
+
v: policy.denyShell.slice(0, 8).join(' | ') + (policy.denyShell.length > 8 ? ' ...' : ''),
|
|
238
|
+
}),
|
|
239
|
+
(0, i18n_1.t)('policy.denyPaths', {
|
|
240
|
+
n: policy.denyPaths.length,
|
|
241
|
+
v: policy.denyPaths.slice(0, 8).join(' | ') + (policy.denyPaths.length > 8 ? ' ...' : ''),
|
|
242
|
+
}),
|
|
236
243
|
'',
|
|
237
|
-
'
|
|
238
|
-
...tenant_1.ROLES.map((r) =>
|
|
244
|
+
(0, i18n_1.t)('policy.matrixTitle'),
|
|
245
|
+
...tenant_1.ROLES.map((r) => (0, i18n_1.t)('policy.matrixRow', {
|
|
246
|
+
role: r.padEnd(10),
|
|
247
|
+
allow: policy.roles[r].allowTools.join(','),
|
|
248
|
+
approval: policy.roles[r].approvalTools.join(',') || '-',
|
|
249
|
+
max: policy.roles[r].maxCostUsd || '∞',
|
|
250
|
+
})),
|
|
239
251
|
];
|
|
240
252
|
return lines.join('\n');
|
|
241
253
|
}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"policy.js","sourceRoot":"","sources":["../../src/enterprise/policy.ts"],"names":[],"mappings":";;;
|
|
1
|
+
{"version":3,"file":"policy.js","sourceRoot":"","sources":["../../src/enterprise/policy.ts"],"names":[],"mappings":";;;AAsKA,gCAaC;AAyBD,4BA4DC;AAGD,oCA8BC;AAzSD;;;;;;;;;;;;;;;GAeG;AACH,2BAA8C;AAC9C,+BAA0C;AAC1C,6CAA0C;AAC1C,qCAA4C;AAC5C,yCAAmC;AAwBtB,QAAA,cAAc,GAAW;IACpC,OAAO,EAAE,CAAC;IACV,KAAK,EAAE;QACL,MAAM,EAAE;YACN,UAAU,EAAE,CAAC,WAAW,EAAE,UAAU,EAAE,MAAM,CAAC;YAC7C,aAAa,EAAE,EAAE;YACjB,UAAU,EAAE,GAAG;SAChB;QACD,SAAS,EAAE;YACT,UAAU,EAAE;gBACV,WAAW;gBACX,UAAU;gBACV,MAAM;gBACN,YAAY;gBACZ,WAAW;gBACX,WAAW;gBACX,aAAa;aACd;YACD,aAAa,EAAE,CAAC,WAAW,CAAC;YAC5B,UAAU,EAAE,CAAC;SACd;QACD,QAAQ,EAAE;YACR,UAAU,EAAE,CAAC,GAAG,CAAC;YACjB,aAAa,EAAE,CAAC,WAAW,CAAC;YAC5B,UAAU,EAAE,CAAC;SACd;QACD,KAAK,EAAE;YACL,UAAU,EAAE,CAAC,GAAG,CAAC;YACjB,aAAa,EAAE,CAAC,WAAW,CAAC;YAC5B,UAAU,EAAE,CAAC;SACd;KACF;IACD,SAAS,EAAE;QACT,UAAU;QACV,UAAU;QACV,UAAU;QACV,MAAM;QACN,QAAQ;QACR,MAAM;QACN,UAAU;QACV,QAAQ;QACR,MAAM;QACN,SAAS;QACT,QAAQ;QACR,OAAO;QACP,aAAa;QACb,eAAe;QACf,WAAW;QACX,SAAS;QACT,WAAW;QACX,SAAS;QACT,aAAa;QACb,UAAU;QACV,YAAY;QACZ,YAAY;QACZ,QAAQ;KACT;IACD,SAAS,EAAE;QACT,MAAM;QACN,aAAa;QACb,YAAY;QACZ,QAAQ;QACR,MAAM;QACN,QAAQ;QACR,YAAY;QACZ,aAAa;QACb,MAAM;QACN,cAAc;QACd,QAAQ;KACT;IACD,oBAAoB,EAAE,CAAC;CACxB,CAAC;AAmBF,qCAAqC;AACrC,SAAS,WAAW,CAAC,IAAY,EAAE,KAA6B;IAC9D,IAAI,CAAC,KAAK;QAAE,OAAO,IAAI,CAAC;IACxB,MAAM,KAAK,GAAG,EAAE,GAAG,IAAI,CAAC,KAAK,EAAE,CAAC;IAChC,KAAK,MAAM,CAAC,IAAI,cAAK,EAAE,CAAC;QACtB,MAAM,CAAC,GAAG,KAAK,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC,CAAC;QAC3B,IAAI,CAAC;YAAE,KAAK,CAAC,CAAC,CAAC,GAAG,EAAE,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,CAAC;IAC/C,CAAC;IACD,OAAO;QACL,OAAO,EAAE,CAAC;QACV,KAAK;QACL,2BAA2B;QAC3B,SAAS,EAAE,CAAC,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,IAAI,CAAC,SAAS,EAAE,GAAG,CAAC,KAAK,CAAC,SAAS,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC;QACxE,SAAS,EAAE,CAAC,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,IAAI,CAAC,SAAS,EAAE,GAAG,CAAC,KAAK,CAAC,SAAS,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC;QACxE,oBAAoB,EAAE,KAAK,CAAC,oBAAoB,IAAI,IAAI,CAAC,oBAAoB;KAC9E,CAAC;AACJ,CAAC;AAED,SAAS,cAAc,CAAC,IAAY;IAClC,IAAI,CAAC,IAAA,eAAU,EAAC,IAAI,CAAC;QAAE,OAAO,IAAI,CAAC;IACnC,IAAI,CAAC;QACH,OAAO,IAAI,CAAC,KAAK,CAAC,IAAA,iBAAY,EAAC,IAAI,EAAE,MAAM,CAAC,CAAoB,CAAC;IACnE,CAAC;IAAC,OAAO,CAAC,EAAE,CAAC;QACX,eAAM,CAAC,IAAI,CAAC,cAAc,EAAE;YAC1B,IAAI;YACJ,KAAK,EAAE,CAAC,YAAY,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC;SAClD,CAAC,CAAC;QACH,OAAO,IAAI,CAAC;IACd,CAAC;AACH,CAAC;AAED,+DAA+D;AAC/D,SAAgB,UAAU,CAAC,QAAgB,EAAE,UAAmB;IAC9D,IAAI,MAAM,GAAG,sBAAc,CAAC;IAC5B,MAAM,GAAG,WAAW,CAAC,MAAM,EAAE,cAAc,CAAC,IAAA,WAAI,EAAC,QAAQ,EAAE,aAAa,CAAC,CAAC,CAAC,CAAC;IAC5E,IAAI,UAAU;QAAE,MAAM,GAAG,WAAW,CAAC,MAAM,EAAE,cAAc,CAAC,IAAA,WAAI,EAAC,UAAU,EAAE,aAAa,CAAC,CAAC,CAAC,CAAC;IAC9F,MAAM,MAAM,GAAG,OAAO,CAAC,GAAG,CAAC,SAAS,EAAE,IAAI,EAAE,CAAC;IAC7C,IAAI,MAAM,EAAE,CAAC;QACX,IAAI,CAAC;YACH,MAAM,GAAG,WAAW,CAAC,MAAM,EAAE,IAAI,CAAC,KAAK,CAAC,MAAM,CAAoB,CAAC,CAAC;QACtE,CAAC;QAAC,MAAM,CAAC;YACP,eAAM,CAAC,IAAI,CAAC,yBAAyB,CAAC,CAAC;QACzC,CAAC;IACH,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,oEAAoE;AACpE,SAAS,YAAY,CAAC,IAA6B;IACjD,MAAM,IAAI,GAAG,CAAC,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,QAAQ,EAAE,MAAM,EAAE,IAAI,CAAC,CAAC;IAC7D,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,KAAK,MAAM,CAAC,IAAI,IAAI,EAAE,CAAC;QACrB,MAAM,CAAC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC;QAClB,IAAI,OAAO,CAAC,KAAK,QAAQ,IAAI,CAAC,CAAC,IAAI,EAAE;YAAE,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;IAC5D,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,SAAS,iBAAiB,CAAC,CAAS;IAClC,OAAO,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,WAAW,EAAE,CAAC;AAC7C,CAAC;AAED,2CAA2C;AAC3C,SAAS,gBAAgB,CAAC,GAAW,EAAE,CAAS;IAC9C,MAAM,GAAG,GAAG,IAAA,cAAO,EAAC,GAAG,EAAE,CAAC,CAAC,CAAC;IAC5B,MAAM,IAAI,GAAG,IAAA,cAAO,EAAC,GAAG,CAAC,CAAC;IAC1B,OAAO,GAAG,KAAK,IAAI,IAAI,CAAC,GAAG,CAAC,UAAU,CAAC,IAAI,GAAG,UAAG,CAAC,CAAC;AACrD,CAAC;AAED,aAAa;AACb,SAAgB,QAAQ,CAAC,MAAc,EAAE,KAAgB;IACvD,MAAM,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,GAAG,EAAE,cAAc,EAAE,GAAG,KAAK,CAAC;IACxD,MAAM,UAAU,GAAG,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAEtC,kCAAkC;IAClC,IAAI,IAAI,KAAK,WAAW,EAAE,CAAC;QACzB,MAAM,GAAG,GAAG,MAAM,CAAC,IAAI,CAAC,OAAO,IAAI,IAAI,CAAC,GAAG,IAAI,EAAE,CAAC,CAAC;QACnD,MAAM,IAAI,GAAG,iBAAiB,CAAC,GAAG,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC;QACzD,KAAK,MAAM,GAAG,IAAI,MAAM,CAAC,SAAS,EAAE,CAAC;YACnC,IAAI,IAAI,CAAC,QAAQ,CAAC,iBAAiB,CAAC,GAAG,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,EAAE,CAAC;gBAC/D,OAAO;oBACL,MAAM,EAAE,MAAM;oBACd,MAAM,EAAE,cAAc,GAAG,EAAE;oBAC3B,IAAI,EAAE,WAAW;iBAClB,CAAC;YACJ,CAAC;QACH,CAAC;IACH,CAAC;IAED,oBAAoB;IACpB,KAAK,MAAM,CAAC,IAAI,YAAY,CAAC,IAAI,CAAC,EAAE,CAAC;QACnC,MAAM,EAAE,GAAG,iBAAiB,CAAC,CAAC,CAAC,CAAC;QAChC,KAAK,MAAM,GAAG,IAAI,MAAM,CAAC,SAAS,EAAE,CAAC;YACnC,MAAM,EAAE,GAAG,iBAAiB,CAAC,GAAG,CAAC,CAAC;YAClC,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,QAAQ,CAAC,GAAG,GAAG,EAAE,CAAC,IAAI,EAAE,CAAC,QAAQ,CAAC,EAAE,GAAG,GAAG,CAAC,IAAI,EAAE,CAAC,QAAQ,CAAC,GAAG,GAAG,EAAE,CAAC,EAAE,CAAC;gBACzF,OAAO,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,cAAc,GAAG,EAAE,EAAE,IAAI,EAAE,WAAW,EAAE,CAAC;YAC5E,CAAC;QACH,CAAC;QACD,IAAI,gBAAgB,CAAC,GAAG,EAAE,CAAC,CAAC,EAAE,CAAC;YAC7B,OAAO,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,YAAY,CAAC,EAAE,EAAE,IAAI,EAAE,SAAS,EAAE,CAAC;QACtE,CAAC;IACH,CAAC;IAED,aAAa;IACb,MAAM,OAAO,GAAG,UAAU,CAAC,UAAU,CAAC,QAAQ,CAAC,GAAG,CAAC,IAAI,UAAU,CAAC,UAAU,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC;IAC5F,MAAM,aAAa,GAAG,UAAU,CAAC,aAAa,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC;IAE9D,IAAI,CAAC,OAAO,IAAI,CAAC,aAAa,EAAE,CAAC;QAC/B,OAAO;YACL,MAAM,EAAE,MAAM;YACd,MAAM,EAAE,MAAM,IAAI,WAAW,IAAI,EAAE;YACnC,IAAI,EAAE,MAAM;SACb,CAAC;IACJ,CAAC;IAED,mBAAmB;IACnB,IAAI,IAAI,KAAK,WAAW,IAAI,aAAa,EAAE,CAAC;QAC1C,MAAM,GAAG,GAAG,MAAM,CAAC,IAAI,CAAC,OAAO,IAAI,IAAI,CAAC,GAAG,IAAI,EAAE,CAAC,CAAC;QACnD,MAAM,IAAI,GAAG,GAAG,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QAC9C,IAAI,cAAc,CAAC,QAAQ,CAAC,IAAI,CAAC,EAAE,CAAC;YAClC,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,iBAAiB,IAAI,EAAE,EAAE,IAAI,EAAE,gBAAgB,EAAE,CAAC;QACtF,CAAC;QACD,OAAO,EAAE,MAAM,EAAE,UAAU,EAAE,MAAM,EAAE,MAAM,IAAI,QAAQ,EAAE,IAAI,EAAE,eAAe,EAAE,CAAC;IACnF,CAAC;IAED,IAAI,aAAa,IAAI,CAAC,OAAO,EAAE,CAAC;QAC9B,OAAO,EAAE,MAAM,EAAE,UAAU,EAAE,MAAM,EAAE,MAAM,IAAI,QAAQ,EAAE,IAAI,EAAE,eAAe,EAAE,CAAC;IACnF,CAAC;IAED,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,MAAM,IAAI,SAAS,IAAI,EAAE,EAAE,IAAI,EAAE,YAAY,EAAE,CAAC;AACpF,CAAC;AAED,4BAA4B;AAC5B,SAAgB,YAAY,CAAC,MAAc,EAAE,IAAU;IACrD,MAAM,EAAE,GAAG,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAC9B,MAAM,KAAK,GAAG;QACZ,IAAA,QAAC,EAAC,gBAAgB,EAAE,EAAE,CAAC,EAAE,MAAM,CAAC,OAAO,EAAE,CAAC;QAC1C,IAAA,QAAC,EAAC,aAAa,EAAE,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC;QAC7B,IAAA,QAAC,EAAC,cAAc,EAAE,EAAE,CAAC,EAAE,EAAE,CAAC,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;QAClD,IAAA,QAAC,EAAC,iBAAiB,EAAE,EAAE,CAAC,EAAE,EAAE,CAAC,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,IAAA,QAAC,EAAC,iBAAiB,CAAC,EAAE,CAAC;QAChF,IAAA,QAAC,EAAC,gBAAgB,EAAE,EAAE,CAAC,EAAE,EAAE,CAAC,UAAU,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,UAAU,EAAE,CAAC,CAAC,CAAC,IAAA,QAAC,EAAC,sBAAsB,CAAC,EAAE,CAAC;QAC/F,IAAA,QAAC,EAAC,kBAAkB,EAAE,EAAE,CAAC,EAAE,MAAM,CAAC,oBAAoB,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,MAAM,CAAC,oBAAoB,EAAE,CAAC,CAAC,CAAC,IAAA,QAAC,EAAC,sBAAsB,CAAC,EAAE,CAAC;QAC7H,IAAA,QAAC,EAAC,kBAAkB,EAAE;YACpB,CAAC,EAAE,MAAM,CAAC,SAAS,CAAC,MAAM;YAC1B,CAAC,EAAE,MAAM,CAAC,SAAS,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,MAAM,CAAC,SAAS,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC;SAC1F,CAAC;QACF,IAAA,QAAC,EAAC,kBAAkB,EAAE;YACpB,CAAC,EAAE,MAAM,CAAC,SAAS,CAAC,MAAM;YAC1B,CAAC,EAAE,MAAM,CAAC,SAAS,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,MAAM,CAAC,SAAS,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC;SAC1F,CAAC;QACF,EAAE;QACF,IAAA,QAAC,EAAC,oBAAoB,CAAC;QACvB,GAAG,cAAK,CAAC,GAAG,CACV,CAAC,CAAC,EAAE,EAAE,CACJ,IAAA,QAAC,EAAC,kBAAkB,EAAE;YACpB,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,EAAE,CAAC;YAClB,KAAK,EAAE,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,IAAI,CAAC,GAAG,CAAC;YAC3C,QAAQ,EAAE,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,GAAG;YACxD,GAAG,EAAE,MAAM,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,UAAU,IAAI,GAAG;SACvC,CAAC,CACL;KACF,CAAC;IACF,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAC1B,CAAC"}
|
|
@@ -0,0 +1,127 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.RealModelExecutor = exports.MockOrchestratorExecutor = exports.SOLUTION_FILE = void 0;
|
|
4
|
+
/**
|
|
5
|
+
* 飞虹 Code (Muse Code 参照复刻)
|
|
6
|
+
* 晋江市飞虹智科技企业管理有限公司 · 飞扬企源研发中心 · 负责人:吴赐虹
|
|
7
|
+
*
|
|
8
|
+
* harness 执行器(可插拔):对单个评测实例运行 Agent 闭环。
|
|
9
|
+
* - HarnessExecutor 接口:mock / 真实模型 / 容器化执行均可实现该接口接入
|
|
10
|
+
* - MockOrchestratorExecutor:用 ScriptedMockProvider 跑「写方案→总结」闭环,
|
|
11
|
+
* 无需模型/网络,验证 pipeline 可用(复刻自 scripts/eval-swebench.mjs runInstance)
|
|
12
|
+
* - RealModelExecutor:用 loadConfig() 构建的真实 ModelRouter(读取 .env 配置,
|
|
13
|
+
* 密钥不经过命令行)驱动编排器,跑真实模型基准(对应 M2.2 真实模型执行)
|
|
14
|
+
*/
|
|
15
|
+
const fs_1 = require("fs");
|
|
16
|
+
const os_1 = require("os");
|
|
17
|
+
const path_1 = require("path");
|
|
18
|
+
const crypto_1 = require("crypto");
|
|
19
|
+
const orchestrator_1 = require("../agent/orchestrator");
|
|
20
|
+
const model_router_1 = require("../models/model-router");
|
|
21
|
+
const mock_provider_1 = require("../models/providers/mock.provider");
|
|
22
|
+
const tools_1 = require("../tools");
|
|
23
|
+
const event_log_1 = require("../runtime/event-log");
|
|
24
|
+
const session_store_1 = require("../runtime/session-store");
|
|
25
|
+
const config_1 = require("../shared/config");
|
|
26
|
+
/** 方案文件名(验证器默认检查该文件是否生成) */
|
|
27
|
+
exports.SOLUTION_FILE = 'SOLUTION.md';
|
|
28
|
+
/** 共享编排循环:临时工作区 + 事件日志 + 会话 + 工具注册表,注入指定 router 跑一轮 */
|
|
29
|
+
async function runOrchestratorLoop(instance, router, maxIterations) {
|
|
30
|
+
const runId = (0, crypto_1.randomUUID)();
|
|
31
|
+
const logDir = (0, fs_1.mkdtempSync)((0, path_1.join)((0, os_1.tmpdir)(), 'fhcode-harness-'));
|
|
32
|
+
const cwd = (0, fs_1.mkdtempSync)((0, path_1.join)((0, os_1.tmpdir)(), 'fhcode-harness-ws-'));
|
|
33
|
+
const stats = { toolCalls: 0 };
|
|
34
|
+
const cleanup = () => {
|
|
35
|
+
(0, fs_1.rmSync)(logDir, { recursive: true, force: true });
|
|
36
|
+
(0, fs_1.rmSync)(cwd, { recursive: true, force: true });
|
|
37
|
+
};
|
|
38
|
+
try {
|
|
39
|
+
const tools = (0, tools_1.createDefaultRegistry)();
|
|
40
|
+
const eventLog = new event_log_1.EventLog(runId, logDir);
|
|
41
|
+
const session = new session_store_1.SessionStore(runId, cwd);
|
|
42
|
+
const orch = new orchestrator_1.Orchestrator({
|
|
43
|
+
router,
|
|
44
|
+
tools,
|
|
45
|
+
eventLog,
|
|
46
|
+
session,
|
|
47
|
+
cwd,
|
|
48
|
+
security: { shellAllowlist: [], requireApproval: false },
|
|
49
|
+
maxIterations,
|
|
50
|
+
maxCostUsd: 0,
|
|
51
|
+
onEvent: (ev) => {
|
|
52
|
+
if (ev.type === 'tool.result')
|
|
53
|
+
stats.toolCalls++;
|
|
54
|
+
},
|
|
55
|
+
});
|
|
56
|
+
const result = await orch.run(instance.problem_statement || instance.instance_id);
|
|
57
|
+
return {
|
|
58
|
+
instance_id: instance.instance_id,
|
|
59
|
+
repo: instance.repo,
|
|
60
|
+
problem: (instance.problem_statement || '').split('\n')[0].slice(0, 80),
|
|
61
|
+
failToPass: instance.FAIL_TO_PASS.length,
|
|
62
|
+
passToPass: instance.PASS_TO_PASS.length,
|
|
63
|
+
runOk: result.ok,
|
|
64
|
+
iterations: result.iterations,
|
|
65
|
+
toolCalls: stats.toolCalls,
|
|
66
|
+
cwd,
|
|
67
|
+
cleanup,
|
|
68
|
+
};
|
|
69
|
+
}
|
|
70
|
+
catch (e) {
|
|
71
|
+
cleanup();
|
|
72
|
+
throw e;
|
|
73
|
+
}
|
|
74
|
+
}
|
|
75
|
+
/** Mock 闭环执行器:无需真实模型/网络,验证 pipeline 可用并记录指标 */
|
|
76
|
+
class MockOrchestratorExecutor {
|
|
77
|
+
opts;
|
|
78
|
+
id = 'mock-orchestrator';
|
|
79
|
+
constructor(opts = {}) {
|
|
80
|
+
this.opts = opts;
|
|
81
|
+
}
|
|
82
|
+
async execute(instance) {
|
|
83
|
+
// 问题陈述作为任务目标;mock 步骤:写方案 → 总结
|
|
84
|
+
const steps = [
|
|
85
|
+
{
|
|
86
|
+
message: {
|
|
87
|
+
role: 'assistant',
|
|
88
|
+
content: '',
|
|
89
|
+
toolCalls: [
|
|
90
|
+
{
|
|
91
|
+
id: 't1',
|
|
92
|
+
name: 'write_file',
|
|
93
|
+
arguments: {
|
|
94
|
+
path: exports.SOLUTION_FILE,
|
|
95
|
+
content: `# Solution for ${instance.instance_id}\n\n${(instance.problem_statement || '').slice(0, 500)}`,
|
|
96
|
+
},
|
|
97
|
+
},
|
|
98
|
+
],
|
|
99
|
+
},
|
|
100
|
+
},
|
|
101
|
+
{ message: { role: 'assistant', content: `已完成 ${instance.instance_id} 方案`, toolCalls: [] } },
|
|
102
|
+
];
|
|
103
|
+
const router = new model_router_1.ModelRouter([new mock_provider_1.ScriptedMockProvider(steps)], 'capability', 0);
|
|
104
|
+
return runOrchestratorLoop(instance, router, this.opts.maxIterations ?? 4);
|
|
105
|
+
}
|
|
106
|
+
}
|
|
107
|
+
exports.MockOrchestratorExecutor = MockOrchestratorExecutor;
|
|
108
|
+
/** 真实模型执行器:读取 .env 配置(loadDotEnv/loadConfig),真实模型驱动编排器跑基准 */
|
|
109
|
+
class RealModelExecutor {
|
|
110
|
+
opts;
|
|
111
|
+
id = 'real-model';
|
|
112
|
+
constructor(opts = {}) {
|
|
113
|
+
this.opts = opts;
|
|
114
|
+
}
|
|
115
|
+
async execute(instance) {
|
|
116
|
+
(0, config_1.loadDotEnv)();
|
|
117
|
+
const cfg = (0, config_1.loadConfig)();
|
|
118
|
+
if (cfg.models.providers.length === 0) {
|
|
119
|
+
throw new Error('未配置任何模型 provider(请设置 FH_PROVIDERS 或 FH_MODEL_NAME)');
|
|
120
|
+
}
|
|
121
|
+
const router = model_router_1.ModelRouter.fromConfig(cfg);
|
|
122
|
+
await router.loadStats(cfg.app.homeDir);
|
|
123
|
+
return runOrchestratorLoop(instance, router, this.opts.maxIterations ?? 10);
|
|
124
|
+
}
|
|
125
|
+
}
|
|
126
|
+
exports.RealModelExecutor = RealModelExecutor;
|
|
127
|
+
//# sourceMappingURL=executor.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"executor.js","sourceRoot":"","sources":["../../src/harness/executor.ts"],"names":[],"mappings":";;;AAAA;;;;;;;;;;GAUG;AACH,2BAAyC;AACzC,2BAA4B;AAC5B,+BAA4B;AAC5B,mCAAoC;AACpC,wDAAqD;AACrD,yDAAqD;AACrD,qEAAwF;AACxF,oCAAiD;AACjD,oDAAgD;AAChD,4DAAwD;AACxD,6CAA0D;AAS1D,4BAA4B;AACf,QAAA,aAAa,GAAG,aAAa,CAAC;AAE3C,uDAAuD;AACvD,KAAK,UAAU,mBAAmB,CAChC,QAAyB,EACzB,MAAmB,EACnB,aAAqB;IAErB,MAAM,KAAK,GAAG,IAAA,mBAAU,GAAE,CAAC;IAC3B,MAAM,MAAM,GAAG,IAAA,gBAAW,EAAC,IAAA,WAAI,EAAC,IAAA,WAAM,GAAE,EAAE,iBAAiB,CAAC,CAAC,CAAC;IAC9D,MAAM,GAAG,GAAG,IAAA,gBAAW,EAAC,IAAA,WAAI,EAAC,IAAA,WAAM,GAAE,EAAE,oBAAoB,CAAC,CAAC,CAAC;IAC9D,MAAM,KAAK,GAAG,EAAE,SAAS,EAAE,CAAC,EAAE,CAAC;IAC/B,MAAM,OAAO,GAAG,GAAS,EAAE;QACzB,IAAA,WAAM,EAAC,MAAM,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;QACjD,IAAA,WAAM,EAAC,GAAG,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;IAChD,CAAC,CAAC;IACF,IAAI,CAAC;QACH,MAAM,KAAK,GAAG,IAAA,6BAAqB,GAAE,CAAC;QACtC,MAAM,QAAQ,GAAG,IAAI,oBAAQ,CAAC,KAAK,EAAE,MAAM,CAAC,CAAC;QAC7C,MAAM,OAAO,GAAG,IAAI,4BAAY,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC;QAC7C,MAAM,IAAI,GAAG,IAAI,2BAAY,CAAC;YAC5B,MAAM;YACN,KAAK;YACL,QAAQ;YACR,OAAO;YACP,GAAG;YACH,QAAQ,EAAE,EAAE,cAAc,EAAE,EAAE,EAAE,eAAe,EAAE,KAAK,EAAE;YACxD,aAAa;YACb,UAAU,EAAE,CAAC;YACb,OAAO,EAAE,CAAC,EAAE,EAAE,EAAE;gBACd,IAAI,EAAE,CAAC,IAAI,KAAK,aAAa;oBAAE,KAAK,CAAC,SAAS,EAAE,CAAC;YACnD,CAAC;SACF,CAAC,CAAC;QACH,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,GAAG,CAAC,QAAQ,CAAC,iBAAiB,IAAI,QAAQ,CAAC,WAAW,CAAC,CAAC;QAClF,OAAO;YACL,WAAW,EAAE,QAAQ,CAAC,WAAW;YACjC,IAAI,EAAE,QAAQ,CAAC,IAAI;YACnB,OAAO,EAAE,CAAC,QAAQ,CAAC,iBAAiB,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC;YACvE,UAAU,EAAE,QAAQ,CAAC,YAAY,CAAC,MAAM;YACxC,UAAU,EAAE,QAAQ,CAAC,YAAY,CAAC,MAAM;YACxC,KAAK,EAAE,MAAM,CAAC,EAAE;YAChB,UAAU,EAAE,MAAM,CAAC,UAAU;YAC7B,SAAS,EAAE,KAAK,CAAC,SAAS;YAC1B,GAAG;YACH,OAAO;SACR,CAAC;IACJ,CAAC;IAAC,OAAO,CAAC,EAAE,CAAC;QACX,OAAO,EAAE,CAAC;QACV,MAAM,CAAC,CAAC;IACV,CAAC;AACH,CAAC;AAED,+CAA+C;AAC/C,MAAa,wBAAwB;IAGN;IAFpB,EAAE,GAAG,mBAAmB,CAAC;IAElC,YAA6B,OAAmC,EAAE;QAArC,SAAI,GAAJ,IAAI,CAAiC;IAAG,CAAC;IAEtE,KAAK,CAAC,OAAO,CAAC,QAAyB;QACrC,8BAA8B;QAC9B,MAAM,KAAK,GAAe;YACxB;gBACE,OAAO,EAAE;oBACP,IAAI,EAAE,WAAW;oBACjB,OAAO,EAAE,EAAE;oBACX,SAAS,EAAE;wBACT;4BACE,EAAE,EAAE,IAAI;4BACR,IAAI,EAAE,YAAY;4BAClB,SAAS,EAAE;gCACT,IAAI,EAAE,qBAAa;gCACnB,OAAO,EAAE,kBAAkB,QAAQ,CAAC,WAAW,OAAO,CAAC,QAAQ,CAAC,iBAAiB,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE;6BACzG;yBACF;qBACF;iBACF;aACF;YACD,EAAE,OAAO,EAAE,EAAE,IAAI,EAAE,WAAW,EAAE,OAAO,EAAE,OAAO,QAAQ,CAAC,WAAW,KAAK,EAAE,SAAS,EAAE,EAAE,EAAE,EAAE;SAC7F,CAAC;QACF,MAAM,MAAM,GAAG,IAAI,0BAAW,CAAC,CAAC,IAAI,oCAAoB,CAAC,KAAK,CAAC,CAAC,EAAE,YAAY,EAAE,CAAC,CAAC,CAAC;QACnF,OAAO,mBAAmB,CAAC,QAAQ,EAAE,MAAM,EAAE,IAAI,CAAC,IAAI,CAAC,aAAa,IAAI,CAAC,CAAC,CAAC;IAC7E,CAAC;CACF;AA7BD,4DA6BC;AAED,6DAA6D;AAC7D,MAAa,iBAAiB;IAGC;IAFpB,EAAE,GAAG,YAAY,CAAC;IAE3B,YAA6B,OAAuD,EAAE;QAAzD,SAAI,GAAJ,IAAI,CAAqD;IAAG,CAAC;IAE1F,KAAK,CAAC,OAAO,CAAC,QAAyB;QACrC,IAAA,mBAAU,GAAE,CAAC;QACb,MAAM,GAAG,GAAG,IAAA,mBAAU,GAAE,CAAC;QACzB,IAAI,GAAG,CAAC,MAAM,CAAC,SAAS,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YACtC,MAAM,IAAI,KAAK,CAAC,oDAAoD,CAAC,CAAC;QACxE,CAAC;QACD,MAAM,MAAM,GAAG,0BAAW,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC;QAC3C,MAAM,MAAM,CAAC,SAAS,CAAC,GAAG,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC;QACxC,OAAO,mBAAmB,CAAC,QAAQ,EAAE,MAAM,EAAE,IAAI,CAAC,IAAI,CAAC,aAAa,IAAI,EAAE,CAAC,CAAC;IAC9E,CAAC;CACF;AAfD,8CAeC"}
|
|
@@ -0,0 +1,87 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
|
|
3
|
+
if (k2 === undefined) k2 = k;
|
|
4
|
+
var desc = Object.getOwnPropertyDescriptor(m, k);
|
|
5
|
+
if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
|
|
6
|
+
desc = { enumerable: true, get: function() { return m[k]; } };
|
|
7
|
+
}
|
|
8
|
+
Object.defineProperty(o, k2, desc);
|
|
9
|
+
}) : (function(o, m, k, k2) {
|
|
10
|
+
if (k2 === undefined) k2 = k;
|
|
11
|
+
o[k2] = m[k];
|
|
12
|
+
}));
|
|
13
|
+
var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
|
|
14
|
+
Object.defineProperty(o, "default", { enumerable: true, value: v });
|
|
15
|
+
}) : function(o, v) {
|
|
16
|
+
o["default"] = v;
|
|
17
|
+
});
|
|
18
|
+
var __importStar = (this && this.__importStar) || (function () {
|
|
19
|
+
var ownKeys = function(o) {
|
|
20
|
+
ownKeys = Object.getOwnPropertyNames || function (o) {
|
|
21
|
+
var ar = [];
|
|
22
|
+
for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
|
|
23
|
+
return ar;
|
|
24
|
+
};
|
|
25
|
+
return ownKeys(o);
|
|
26
|
+
};
|
|
27
|
+
return function (mod) {
|
|
28
|
+
if (mod && mod.__esModule) return mod;
|
|
29
|
+
var result = {};
|
|
30
|
+
if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
|
|
31
|
+
__setModuleDefault(result, mod);
|
|
32
|
+
return result;
|
|
33
|
+
};
|
|
34
|
+
})();
|
|
35
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
36
|
+
exports.Harness = void 0;
|
|
37
|
+
/** harness 编排器:load → execute → verify → report */
|
|
38
|
+
class Harness {
|
|
39
|
+
opts;
|
|
40
|
+
constructor(opts) {
|
|
41
|
+
this.opts = opts;
|
|
42
|
+
}
|
|
43
|
+
async run() {
|
|
44
|
+
const { loader, executor, verifier, reporter, limit = 10, offset = 0, onProgress } = this.opts;
|
|
45
|
+
const all = await loader.load();
|
|
46
|
+
const slice = all.slice(offset, offset + limit);
|
|
47
|
+
const results = [];
|
|
48
|
+
for (const inst of slice) {
|
|
49
|
+
const exec = await executor.execute(inst);
|
|
50
|
+
const verified = verifier ? await verifier.verify(exec.cwd, inst) : true;
|
|
51
|
+
const result = {
|
|
52
|
+
instance_id: exec.instance_id,
|
|
53
|
+
repo: exec.repo,
|
|
54
|
+
problem: exec.problem,
|
|
55
|
+
failToPass: exec.failToPass,
|
|
56
|
+
passToPass: exec.passToPass,
|
|
57
|
+
ok: exec.runOk && verified,
|
|
58
|
+
iterations: exec.iterations,
|
|
59
|
+
toolCalls: exec.toolCalls,
|
|
60
|
+
verified,
|
|
61
|
+
};
|
|
62
|
+
exec.cleanup();
|
|
63
|
+
results.push(result);
|
|
64
|
+
onProgress?.(result, results.length, slice.length);
|
|
65
|
+
}
|
|
66
|
+
const completed = results.filter((r) => r.ok).length;
|
|
67
|
+
const report = {
|
|
68
|
+
meta: {
|
|
69
|
+
split: loader.opts?.split ?? loader.id,
|
|
70
|
+
limit,
|
|
71
|
+
offset,
|
|
72
|
+
mode: executor.id,
|
|
73
|
+
startedAt: new Date().toISOString(),
|
|
74
|
+
},
|
|
75
|
+
results,
|
|
76
|
+
summary: {
|
|
77
|
+
total: results.length,
|
|
78
|
+
completed,
|
|
79
|
+
rate: results.length ? completed / results.length : 0,
|
|
80
|
+
},
|
|
81
|
+
};
|
|
82
|
+
const renderer = reporter ?? new (await Promise.resolve().then(() => __importStar(require('./reporter')))).MarkdownReporter();
|
|
83
|
+
return { report, rendered: renderer.render(report) };
|
|
84
|
+
}
|
|
85
|
+
}
|
|
86
|
+
exports.Harness = Harness;
|
|
87
|
+
//# sourceMappingURL=harness.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"harness.js","sourceRoot":"","sources":["../../src/harness/harness.ts"],"names":[],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AA+BA,mDAAmD;AACnD,MAAa,OAAO;IACW;IAA7B,YAA6B,IAAoB;QAApB,SAAI,GAAJ,IAAI,CAAgB;IAAG,CAAC;IAErD,KAAK,CAAC,GAAG;QACP,MAAM,EAAE,MAAM,EAAE,QAAQ,EAAE,QAAQ,EAAE,QAAQ,EAAE,KAAK,GAAG,EAAE,EAAE,MAAM,GAAG,CAAC,EAAE,UAAU,EAAE,GAAG,IAAI,CAAC,IAAI,CAAC;QAE/F,MAAM,GAAG,GAAG,MAAM,MAAM,CAAC,IAAI,EAAE,CAAC;QAChC,MAAM,KAAK,GAAG,GAAG,CAAC,KAAK,CAAC,MAAM,EAAE,MAAM,GAAG,KAAK,CAAC,CAAC;QAEhD,MAAM,OAAO,GAAoB,EAAE,CAAC;QACpC,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;YACzB,MAAM,IAAI,GAAG,MAAM,QAAQ,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC;YAC1C,MAAM,QAAQ,GAAG,QAAQ,CAAC,CAAC,CAAC,MAAM,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;YACzE,MAAM,MAAM,GAAkB;gBAC5B,WAAW,EAAE,IAAI,CAAC,WAAW;gBAC7B,IAAI,EAAE,IAAI,CAAC,IAAI;gBACf,OAAO,EAAE,IAAI,CAAC,OAAO;gBACrB,UAAU,EAAE,IAAI,CAAC,UAAU;gBAC3B,UAAU,EAAE,IAAI,CAAC,UAAU;gBAC3B,EAAE,EAAE,IAAI,CAAC,KAAK,IAAI,QAAQ;gBAC1B,UAAU,EAAE,IAAI,CAAC,UAAU;gBAC3B,SAAS,EAAE,IAAI,CAAC,SAAS;gBACzB,QAAQ;aACT,CAAC;YACF,IAAI,CAAC,OAAO,EAAE,CAAC;YACf,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YACrB,UAAU,EAAE,CAAC,MAAM,EAAE,OAAO,CAAC,MAAM,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC;QACrD,CAAC;QAED,MAAM,SAAS,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,MAAM,CAAC;QACrD,MAAM,MAAM,GAAkB;YAC5B,IAAI,EAAE;gBACJ,KAAK,EAAG,MAAwC,CAAC,IAAI,EAAE,KAAK,IAAI,MAAM,CAAC,EAAE;gBACzE,KAAK;gBACL,MAAM;gBACN,IAAI,EAAE,QAAQ,CAAC,EAAE;gBACjB,SAAS,EAAE,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE;aACpC;YACD,OAAO;YACP,OAAO,EAAE;gBACP,KAAK,EAAE,OAAO,CAAC,MAAM;gBACrB,SAAS;gBACT,IAAI,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,SAAS,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;aACtD;SACF,CAAC;QAEF,MAAM,QAAQ,GAAG,QAAQ,IAAI,IAAI,CAAC,wDAAa,YAAY,GAAC,CAAC,CAAC,gBAAgB,EAAE,CAAC;QACjF,OAAO,EAAE,MAAM,EAAE,QAAQ,EAAE,QAAQ,CAAC,MAAM,CAAC,MAAM,CAAC,EAAE,CAAC;IACvD,CAAC;CACF;AAjDD,0BAiDC"}
|
|
@@ -0,0 +1,31 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
|
|
3
|
+
if (k2 === undefined) k2 = k;
|
|
4
|
+
var desc = Object.getOwnPropertyDescriptor(m, k);
|
|
5
|
+
if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
|
|
6
|
+
desc = { enumerable: true, get: function() { return m[k]; } };
|
|
7
|
+
}
|
|
8
|
+
Object.defineProperty(o, k2, desc);
|
|
9
|
+
}) : (function(o, m, k, k2) {
|
|
10
|
+
if (k2 === undefined) k2 = k;
|
|
11
|
+
o[k2] = m[k];
|
|
12
|
+
}));
|
|
13
|
+
var __exportStar = (this && this.__exportStar) || function(m, exports) {
|
|
14
|
+
for (var p in m) if (p !== "default" && !Object.prototype.hasOwnProperty.call(exports, p)) __createBinding(exports, m, p);
|
|
15
|
+
};
|
|
16
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
17
|
+
/**
|
|
18
|
+
* 飞虹 Code (Muse Code 参照复刻)
|
|
19
|
+
* 晋江市飞虹智科技企业管理有限公司 · 飞扬企源研发中心 · 负责人:吴赐虹
|
|
20
|
+
*
|
|
21
|
+
* 可插拔 harness 模块组统一出口:
|
|
22
|
+
* - 评测编码/修复能力的完整闭环(加载实例 → 执行 → 验证 → 报告)
|
|
23
|
+
* - 加载器 / 执行器 / 验证器 / 报告器四类组件均可插拔替换
|
|
24
|
+
*/
|
|
25
|
+
__exportStar(require("./types"), exports);
|
|
26
|
+
__exportStar(require("./loader"), exports);
|
|
27
|
+
__exportStar(require("./executor"), exports);
|
|
28
|
+
__exportStar(require("./verifier"), exports);
|
|
29
|
+
__exportStar(require("./reporter"), exports);
|
|
30
|
+
__exportStar(require("./harness"), exports);
|
|
31
|
+
//# sourceMappingURL=index.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"index.js","sourceRoot":"","sources":["../../src/harness/index.ts"],"names":[],"mappings":";;;;;;;;;;;;;;;;AAAA;;;;;;;GAOG;AACH,0CAAwB;AACxB,2CAAyB;AACzB,6CAA2B;AAC3B,6CAA2B;AAC3B,6CAA2B;AAC3B,4CAA0B"}
|
|
@@ -0,0 +1,138 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.LocalJsonLoader = exports.SwebenchLoader = void 0;
|
|
4
|
+
exports.normalizeInstance = normalizeInstance;
|
|
5
|
+
/**
|
|
6
|
+
* 飞虹 Code (Muse Code 参照复刻)
|
|
7
|
+
* 晋江市飞虹智科技企业管理有限公司 · 飞扬企源研发中心 · 负责人:吴赐虹
|
|
8
|
+
*
|
|
9
|
+
* harness 数据集加载器(可插拔):
|
|
10
|
+
* - DatasetLoader 接口:任何数据集(SWE-bench / 本地 JSON / 镜像)实现该接口即可接入
|
|
11
|
+
* - SwebenchLoader:从 HuggingFace datasets-server 分页拉取(默认)或镜像/离线 JSON 加载,
|
|
12
|
+
* 缓存到 ~/.feihong-code/bench/swebench-<split>.json(复刻自 scripts/eval-swebench.mjs)
|
|
13
|
+
* - LocalJsonLoader:加载任意本地/远程 JSON(数组或 { instances: [] })
|
|
14
|
+
*/
|
|
15
|
+
const fs_1 = require("fs");
|
|
16
|
+
const os_1 = require("os");
|
|
17
|
+
const path_1 = require("path");
|
|
18
|
+
const HF_BASE = 'https://datasets-server.huggingface.co';
|
|
19
|
+
const DATASETS = {
|
|
20
|
+
lite: { id: 'SWE-bench/SWE-bench_Lite', split: 'test' },
|
|
21
|
+
verified: { id: 'SWE-bench/SWE-bench_Verified', split: 'test' },
|
|
22
|
+
};
|
|
23
|
+
/** 归一化实例字段(容忍 HF 字符串化数组 / 缺失字段) */
|
|
24
|
+
function normalizeInstance(row) {
|
|
25
|
+
const parseList = (v) => {
|
|
26
|
+
if (Array.isArray(v))
|
|
27
|
+
return v.map(String);
|
|
28
|
+
if (typeof v === 'string' && v.trim()) {
|
|
29
|
+
try {
|
|
30
|
+
const parsed = JSON.parse(v);
|
|
31
|
+
return Array.isArray(parsed) ? parsed.map(String) : [v];
|
|
32
|
+
}
|
|
33
|
+
catch {
|
|
34
|
+
return [v];
|
|
35
|
+
}
|
|
36
|
+
}
|
|
37
|
+
return [];
|
|
38
|
+
};
|
|
39
|
+
const str = (v) => (typeof v === 'string' ? v : '');
|
|
40
|
+
return {
|
|
41
|
+
instance_id: str(row.instance_id) || 'unknown',
|
|
42
|
+
repo: str(row.repo),
|
|
43
|
+
base_commit: str(row.base_commit),
|
|
44
|
+
problem_statement: str(row.problem_statement),
|
|
45
|
+
patch: str(row.patch),
|
|
46
|
+
test_patch: str(row.test_patch),
|
|
47
|
+
FAIL_TO_PASS: parseList(row.FAIL_TO_PASS),
|
|
48
|
+
PASS_TO_PASS: parseList(row.PASS_TO_PASS),
|
|
49
|
+
created_at: str(row.created_at),
|
|
50
|
+
version: str(row.version),
|
|
51
|
+
};
|
|
52
|
+
}
|
|
53
|
+
function cachePath(split) {
|
|
54
|
+
const dir = (0, path_1.join)((0, os_1.homedir)(), '.feihong-code', 'bench');
|
|
55
|
+
return { dir, file: (0, path_1.join)(dir, `swebench-${split}.json`) };
|
|
56
|
+
}
|
|
57
|
+
/** 从 HF datasets-server 分页拉取 rows */
|
|
58
|
+
async function fetchHfRows(dataset, split, offset, length) {
|
|
59
|
+
const url = `${HF_BASE}/rows?dataset=${encodeURIComponent(dataset)}&config=default&split=${split}&offset=${offset}&length=${length}`;
|
|
60
|
+
const res = await fetch(url, { headers: { 'User-Agent': 'fhcode/0.5' } });
|
|
61
|
+
if (!res.ok)
|
|
62
|
+
throw new Error(`HF API HTTP ${res.status}(${url})`);
|
|
63
|
+
const data = (await res.json());
|
|
64
|
+
if (!Array.isArray(data.rows))
|
|
65
|
+
throw new Error('HF API 响应缺少 rows 字段');
|
|
66
|
+
return data.rows.map((r) => r.row);
|
|
67
|
+
}
|
|
68
|
+
/** 从镜像/离线 JSON 文件加载全部实例(数组或 {instances: []} 均可) */
|
|
69
|
+
async function loadMirror(fileOrUrl) {
|
|
70
|
+
const text = /^https?:\/\//i.test(fileOrUrl)
|
|
71
|
+
? await (await fetch(fileOrUrl, { headers: { 'User-Agent': 'fhcode/0.5' } })).text()
|
|
72
|
+
: (0, fs_1.readFileSync)(fileOrUrl, 'utf8');
|
|
73
|
+
const parsed = JSON.parse(text);
|
|
74
|
+
return Array.isArray(parsed) ? parsed : (parsed.instances ?? []);
|
|
75
|
+
}
|
|
76
|
+
/** SWE-bench 数据集加载器(HF / 镜像 / 缓存) */
|
|
77
|
+
class SwebenchLoader {
|
|
78
|
+
opts;
|
|
79
|
+
id = 'swebench';
|
|
80
|
+
constructor(opts = {}) {
|
|
81
|
+
this.opts = opts;
|
|
82
|
+
}
|
|
83
|
+
async load() {
|
|
84
|
+
const split = this.opts.split ?? 'lite';
|
|
85
|
+
const { dir, file } = cachePath(split);
|
|
86
|
+
if (!this.opts.force && (0, fs_1.existsSync)(file)) {
|
|
87
|
+
try {
|
|
88
|
+
const rows = JSON.parse((0, fs_1.readFileSync)(file, 'utf8'));
|
|
89
|
+
if (Array.isArray(rows))
|
|
90
|
+
return rows.map(normalizeInstance);
|
|
91
|
+
}
|
|
92
|
+
catch {
|
|
93
|
+
/* 缓存损坏则重新拉取 */
|
|
94
|
+
}
|
|
95
|
+
}
|
|
96
|
+
const mirror = process.env.FH_SWEBENCH_DATA_URL;
|
|
97
|
+
let rows;
|
|
98
|
+
if (mirror) {
|
|
99
|
+
rows = await loadMirror(mirror);
|
|
100
|
+
}
|
|
101
|
+
else {
|
|
102
|
+
const ds = DATASETS[split];
|
|
103
|
+
if (!ds)
|
|
104
|
+
throw new Error(`未知 split: ${split}(可选 lite|verified)`);
|
|
105
|
+
rows = [];
|
|
106
|
+
for (let offset = 0; offset < 300; offset += 100) {
|
|
107
|
+
const page = await fetchHfRows(ds.id, ds.split, offset, 100);
|
|
108
|
+
rows.push(...page);
|
|
109
|
+
if (page.length < 100)
|
|
110
|
+
break;
|
|
111
|
+
}
|
|
112
|
+
}
|
|
113
|
+
const normalized = rows.map(normalizeInstance);
|
|
114
|
+
try {
|
|
115
|
+
(0, fs_1.mkdirSync)(dir, { recursive: true });
|
|
116
|
+
(0, fs_1.writeFileSync)(file, JSON.stringify(normalized), 'utf8');
|
|
117
|
+
}
|
|
118
|
+
catch {
|
|
119
|
+
/* 缓存失败不影响使用 */
|
|
120
|
+
}
|
|
121
|
+
return normalized;
|
|
122
|
+
}
|
|
123
|
+
}
|
|
124
|
+
exports.SwebenchLoader = SwebenchLoader;
|
|
125
|
+
/** 本地/远程 JSON 数据集加载器(数组或 {instances: []}),便于离线与自定义评测集 */
|
|
126
|
+
class LocalJsonLoader {
|
|
127
|
+
source;
|
|
128
|
+
id = 'local-json';
|
|
129
|
+
constructor(source) {
|
|
130
|
+
this.source = source;
|
|
131
|
+
}
|
|
132
|
+
async load() {
|
|
133
|
+
const rows = await loadMirror(this.source);
|
|
134
|
+
return rows.map(normalizeInstance);
|
|
135
|
+
}
|
|
136
|
+
}
|
|
137
|
+
exports.LocalJsonLoader = LocalJsonLoader;
|
|
138
|
+
//# sourceMappingURL=loader.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"loader.js","sourceRoot":"","sources":["../../src/harness/loader.ts"],"names":[],"mappings":";;;AA4BA,8CA0BC;AAtDD;;;;;;;;;GASG;AACH,2BAAwE;AACxE,2BAA6B;AAC7B,+BAA4B;AAS5B,MAAM,OAAO,GAAG,wCAAwC,CAAC;AACzD,MAAM,QAAQ,GAAkD;IAC9D,IAAI,EAAE,EAAE,EAAE,EAAE,0BAA0B,EAAE,KAAK,EAAE,MAAM,EAAE;IACvD,QAAQ,EAAE,EAAE,EAAE,EAAE,8BAA8B,EAAE,KAAK,EAAE,MAAM,EAAE;CAChE,CAAC;AAEF,mCAAmC;AACnC,SAAgB,iBAAiB,CAAC,GAA4B;IAC5D,MAAM,SAAS,GAAG,CAAC,CAAU,EAAY,EAAE;QACzC,IAAI,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC;YAAE,OAAO,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;QAC3C,IAAI,OAAO,CAAC,KAAK,QAAQ,IAAI,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC;YACtC,IAAI,CAAC;gBACH,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,CAAY,CAAC;gBACxC,OAAO,KAAK,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YAC1D,CAAC;YAAC,MAAM,CAAC;gBACP,OAAO,CAAC,CAAC,CAAC,CAAC;YACb,CAAC;QACH,CAAC;QACD,OAAO,EAAE,CAAC;IACZ,CAAC,CAAC;IACF,MAAM,GAAG,GAAG,CAAC,CAAU,EAAU,EAAE,CAAC,CAAC,OAAO,CAAC,KAAK,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;IACrE,OAAO;QACL,WAAW,EAAE,GAAG,CAAC,GAAG,CAAC,WAAW,CAAC,IAAI,SAAS;QAC9C,IAAI,EAAE,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC;QACnB,WAAW,EAAE,GAAG,CAAC,GAAG,CAAC,WAAW,CAAC;QACjC,iBAAiB,EAAE,GAAG,CAAC,GAAG,CAAC,iBAAiB,CAAC;QAC7C,KAAK,EAAE,GAAG,CAAC,GAAG,CAAC,KAAK,CAAC;QACrB,UAAU,EAAE,GAAG,CAAC,GAAG,CAAC,UAAU,CAAC;QAC/B,YAAY,EAAE,SAAS,CAAC,GAAG,CAAC,YAAY,CAAC;QACzC,YAAY,EAAE,SAAS,CAAC,GAAG,CAAC,YAAY,CAAC;QACzC,UAAU,EAAE,GAAG,CAAC,GAAG,CAAC,UAAU,CAAC;QAC/B,OAAO,EAAE,GAAG,CAAC,GAAG,CAAC,OAAO,CAAC;KAC1B,CAAC;AACJ,CAAC;AAED,SAAS,SAAS,CAAC,KAAa;IAC9B,MAAM,GAAG,GAAG,IAAA,WAAI,EAAC,IAAA,YAAO,GAAE,EAAE,eAAe,EAAE,OAAO,CAAC,CAAC;IACtD,OAAO,EAAE,GAAG,EAAE,IAAI,EAAE,IAAA,WAAI,EAAC,GAAG,EAAE,YAAY,KAAK,OAAO,CAAC,EAAE,CAAC;AAC5D,CAAC;AAED,qCAAqC;AACrC,KAAK,UAAU,WAAW,CAAC,OAAe,EAAE,KAAa,EAAE,MAAc,EAAE,MAAc;IACvF,MAAM,GAAG,GAAG,GAAG,OAAO,iBAAiB,kBAAkB,CAAC,OAAO,CAAC,yBAAyB,KAAK,WAAW,MAAM,WAAW,MAAM,EAAE,CAAC;IACrI,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,GAAG,EAAE,EAAE,OAAO,EAAE,EAAE,YAAY,EAAE,YAAY,EAAE,EAAE,CAAC,CAAC;IAC1E,IAAI,CAAC,GAAG,CAAC,EAAE;QAAE,MAAM,IAAI,KAAK,CAAC,eAAe,GAAG,CAAC,MAAM,IAAI,GAAG,GAAG,CAAC,CAAC;IAClE,MAAM,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC,IAAI,EAAE,CAAkD,CAAC;IACjF,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;QAAE,MAAM,IAAI,KAAK,CAAC,qBAAqB,CAAC,CAAC;IACtE,OAAO,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;AACrC,CAAC;AAED,mDAAmD;AACnD,KAAK,UAAU,UAAU,CAAC,SAAiB;IACzC,MAAM,IAAI,GAAG,eAAe,CAAC,IAAI,CAAC,SAAS,CAAC;QAC1C,CAAC,CAAC,MAAM,CAAC,MAAM,KAAK,CAAC,SAAS,EAAE,EAAE,OAAO,EAAE,EAAE,YAAY,EAAE,YAAY,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE;QACpF,CAAC,CAAC,IAAA,iBAAY,EAAC,SAAS,EAAE,MAAM,CAAC,CAAC;IACpC,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAY,CAAC;IAC3C,OAAO,KAAK,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,CAAE,MAAoC,CAAC,CAAC,CAAC,CAAE,MAAoD,CAAC,SAAS,IAAI,EAAE,CAAC,CAAC;AACjJ,CAAC;AAED,qCAAqC;AACrC,MAAa,cAAc;IAGI;IAFpB,EAAE,GAAG,UAAU,CAAC;IAEzB,YAA6B,OAA4C,EAAE;QAA9C,SAAI,GAAJ,IAAI,CAA0C;IAAG,CAAC;IAE/E,KAAK,CAAC,IAAI;QACR,MAAM,KAAK,GAAG,IAAI,CAAC,IAAI,CAAC,KAAK,IAAI,MAAM,CAAC;QACxC,MAAM,EAAE,GAAG,EAAE,IAAI,EAAE,GAAG,SAAS,CAAC,KAAK,CAAC,CAAC;QACvC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,IAAI,IAAA,eAAU,EAAC,IAAI,CAAC,EAAE,CAAC;YACzC,IAAI,CAAC;gBACH,MAAM,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,IAAA,iBAAY,EAAC,IAAI,EAAE,MAAM,CAAC,CAA8B,CAAC;gBACjF,IAAI,KAAK,CAAC,OAAO,CAAC,IAAI,CAAC;oBAAE,OAAO,IAAI,CAAC,GAAG,CAAC,iBAAiB,CAAC,CAAC;YAC9D,CAAC;YAAC,MAAM,CAAC;gBACP,eAAe;YACjB,CAAC;QACH,CAAC;QACD,MAAM,MAAM,GAAG,OAAO,CAAC,GAAG,CAAC,oBAAoB,CAAC;QAChD,IAAI,IAA+B,CAAC;QACpC,IAAI,MAAM,EAAE,CAAC;YACX,IAAI,GAAG,MAAM,UAAU,CAAC,MAAM,CAAC,CAAC;QAClC,CAAC;aAAM,CAAC;YACN,MAAM,EAAE,GAAG,QAAQ,CAAC,KAAK,CAAC,CAAC;YAC3B,IAAI,CAAC,EAAE;gBAAE,MAAM,IAAI,KAAK,CAAC,aAAa,KAAK,oBAAoB,CAAC,CAAC;YACjE,IAAI,GAAG,EAAE,CAAC;YACV,KAAK,IAAI,MAAM,GAAG,CAAC,EAAE,MAAM,GAAG,GAAG,EAAE,MAAM,IAAI,GAAG,EAAE,CAAC;gBACjD,MAAM,IAAI,GAAG,MAAM,WAAW,CAAC,EAAE,CAAC,EAAE,EAAE,EAAE,CAAC,KAAK,EAAE,MAAM,EAAE,GAAG,CAAC,CAAC;gBAC7D,IAAI,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC,CAAC;gBACnB,IAAI,IAAI,CAAC,MAAM,GAAG,GAAG;oBAAE,MAAM;YAC/B,CAAC;QACH,CAAC;QACD,MAAM,UAAU,GAAG,IAAI,CAAC,GAAG,CAAC,iBAAiB,CAAC,CAAC;QAC/C,IAAI,CAAC;YACH,IAAA,cAAS,EAAC,GAAG,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;YACpC,IAAA,kBAAa,EAAC,IAAI,EAAE,IAAI,CAAC,SAAS,CAAC,UAAU,CAAC,EAAE,MAAM,CAAC,CAAC;QAC1D,CAAC;QAAC,MAAM,CAAC;YACP,eAAe;QACjB,CAAC;QACD,OAAO,UAAU,CAAC;IACpB,CAAC;CACF;AAvCD,wCAuCC;AAED,yDAAyD;AACzD,MAAa,eAAe;IAGG;IAFpB,EAAE,GAAG,YAAY,CAAC;IAE3B,YAA6B,MAAc;QAAd,WAAM,GAAN,MAAM,CAAQ;IAAG,CAAC;IAE/C,KAAK,CAAC,IAAI;QACR,MAAM,IAAI,GAAG,MAAM,UAAU,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAC3C,OAAO,IAAI,CAAC,GAAG,CAAC,iBAAiB,CAAC,CAAC;IACrC,CAAC;CACF;AATD,0CASC"}
|
|
@@ -0,0 +1,34 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.JsonReporter = exports.MarkdownReporter = void 0;
|
|
4
|
+
/** Markdown 报告器:表格 + 汇总(人类可读,可写文件) */
|
|
5
|
+
class MarkdownReporter {
|
|
6
|
+
id = 'markdown';
|
|
7
|
+
render(report) {
|
|
8
|
+
const { meta, results, summary } = report;
|
|
9
|
+
const lines = [];
|
|
10
|
+
lines.push('# SWE-bench 跑分报告');
|
|
11
|
+
lines.push('');
|
|
12
|
+
lines.push(`- 数据集: ${meta.split} · 实例数: ${summary.total} · 完成: ${summary.completed} · 通过率: ${Math.round(summary.rate * 100)}%`);
|
|
13
|
+
lines.push(`- 运行时间: ${meta.startedAt} · 模式: ${meta.mode}`);
|
|
14
|
+
lines.push('');
|
|
15
|
+
lines.push('| instance_id | repo | 问题 | F2P | P2P | 结果 | 迭代 | 工具 |');
|
|
16
|
+
lines.push('|---|---|---|---|---|---|---|---|');
|
|
17
|
+
for (const r of results) {
|
|
18
|
+
lines.push(`| ${r.instance_id} | ${r.repo} | ${r.problem} | ${r.failToPass} | ${r.passToPass} | ${r.ok ? '✅' : '❌'} | ${r.iterations} | ${r.toolCalls} |`);
|
|
19
|
+
}
|
|
20
|
+
lines.push('');
|
|
21
|
+
lines.push(`汇总: ${summary.completed}/${summary.total} 通过(${meta.mode})`);
|
|
22
|
+
return lines.join('\n');
|
|
23
|
+
}
|
|
24
|
+
}
|
|
25
|
+
exports.MarkdownReporter = MarkdownReporter;
|
|
26
|
+
/** JSON 报告器:结构化输出(CI 门禁 / 程序消费) */
|
|
27
|
+
class JsonReporter {
|
|
28
|
+
id = 'json';
|
|
29
|
+
render(report) {
|
|
30
|
+
return JSON.stringify(report, null, 2);
|
|
31
|
+
}
|
|
32
|
+
}
|
|
33
|
+
exports.JsonReporter = JsonReporter;
|
|
34
|
+
//# sourceMappingURL=reporter.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"reporter.js","sourceRoot":"","sources":["../../src/harness/reporter.ts"],"names":[],"mappings":";;;AAeA,sCAAsC;AACtC,MAAa,gBAAgB;IAClB,EAAE,GAAG,UAAU,CAAC;IAEzB,MAAM,CAAC,MAAqB;QAC1B,MAAM,EAAE,IAAI,EAAE,OAAO,EAAE,OAAO,EAAE,GAAG,MAAM,CAAC;QAC1C,MAAM,KAAK,GAAa,EAAE,CAAC;QAC3B,KAAK,CAAC,IAAI,CAAC,kBAAkB,CAAC,CAAC;QAC/B,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACf,KAAK,CAAC,IAAI,CAAC,UAAU,IAAI,CAAC,KAAK,WAAW,OAAO,CAAC,KAAK,UAAU,OAAO,CAAC,SAAS,WAAW,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,IAAI,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC;QAChI,KAAK,CAAC,IAAI,CAAC,WAAW,IAAI,CAAC,SAAS,UAAU,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC;QAC3D,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACf,KAAK,CAAC,IAAI,CAAC,wDAAwD,CAAC,CAAC;QACrE,KAAK,CAAC,IAAI,CAAC,mCAAmC,CAAC,CAAC;QAChD,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;YACxB,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,WAAW,MAAM,CAAC,CAAC,IAAI,MAAM,CAAC,CAAC,OAAO,MAAM,CAAC,CAAC,UAAU,MAAM,CAAC,CAAC,UAAU,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,MAAM,CAAC,CAAC,UAAU,MAAM,CAAC,CAAC,SAAS,IAAI,CAAC,CAAC;QAC7J,CAAC;QACD,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACf,KAAK,CAAC,IAAI,CAAC,OAAO,OAAO,CAAC,SAAS,IAAI,OAAO,CAAC,KAAK,OAAO,IAAI,CAAC,IAAI,GAAG,CAAC,CAAC;QACzE,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IAC1B,CAAC;CACF;AApBD,4CAoBC;AAED,mCAAmC;AACnC,MAAa,YAAY;IACd,EAAE,GAAG,MAAM,CAAC;IAErB,MAAM,CAAC,MAAqB;QAC1B,OAAO,IAAI,CAAC,SAAS,CAAC,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC;IACzC,CAAC;CACF;AAND,oCAMC"}
|
|
@@ -0,0 +1,10 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
/**
|
|
3
|
+
* 飞虹 Code (Muse Code 参照复刻)
|
|
4
|
+
* 晋江市飞虹智科技企业管理有限公司 · 飞扬企源研发中心 · 负责人:吴赐虹
|
|
5
|
+
*
|
|
6
|
+
* harness 模块组核心类型:评测实例 / 执行产出 / 验证结果 / 报告。
|
|
7
|
+
* 对齐 SWE-bench 官方实例结构,供加载器、执行器、验证器、报告器共享。
|
|
8
|
+
*/
|
|
9
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
10
|
+
//# sourceMappingURL=types.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"types.js","sourceRoot":"","sources":["../../src/harness/types.ts"],"names":[],"mappings":";AAAA;;;;;;GAMG"}
|
|
@@ -0,0 +1,48 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.TestVerifier = exports.FileExistsVerifier = void 0;
|
|
4
|
+
/**
|
|
5
|
+
* 飞虹 Code (Muse Code 参照复刻)
|
|
6
|
+
* 晋江市飞虹智科技企业管理有限公司 · 飞扬企源研发中心 · 负责人:吴赐虹
|
|
7
|
+
*
|
|
8
|
+
* harness 验证器(可插拔):判定执行产出是否满足通过条件。
|
|
9
|
+
* - HarnessVerifier 接口:文件存在 / 测试通过 / 自定义规则均可实现接入(异步,跑测试需要)
|
|
10
|
+
* - FileExistsVerifier:检查工作区是否生成了指定文件(当前 mock 闭环的通过标准)
|
|
11
|
+
* - TestVerifier:在工作区运行测试命令判定通过——实例带 FAIL_TO_PASS 时
|
|
12
|
+
* 优先以这些失败用例为测试目标(`npm test -- <用例>`),否则跑默认测试套件。
|
|
13
|
+
* 命令经 sanitizeManagedCommand 约束(仅包管理器脚本,防命令注入)。
|
|
14
|
+
*/
|
|
15
|
+
const fs_1 = require("fs");
|
|
16
|
+
const path_1 = require("path");
|
|
17
|
+
const exec_1 = require("../tools/shell/exec");
|
|
18
|
+
/** 文件存在验证器:默认检查方案文件是否生成 */
|
|
19
|
+
class FileExistsVerifier {
|
|
20
|
+
file;
|
|
21
|
+
id = 'file-exists';
|
|
22
|
+
constructor(file = 'SOLUTION.md') {
|
|
23
|
+
this.file = file;
|
|
24
|
+
}
|
|
25
|
+
async verify(cwd, _instance) {
|
|
26
|
+
return (0, fs_1.existsSync)((0, path_1.join)(cwd, this.file));
|
|
27
|
+
}
|
|
28
|
+
}
|
|
29
|
+
exports.FileExistsVerifier = FileExistsVerifier;
|
|
30
|
+
/** 测试通过型验证器:在工作区运行测试,退出码 0 即通过 */
|
|
31
|
+
class TestVerifier {
|
|
32
|
+
opts;
|
|
33
|
+
id = 'test-run';
|
|
34
|
+
constructor(opts = {}) {
|
|
35
|
+
this.opts = opts;
|
|
36
|
+
}
|
|
37
|
+
async verify(cwd, instance) {
|
|
38
|
+
// 有 FAIL_TO_PASS 用例时优先跑这些用例(回归到「失败用例现在应通过」);否则跑默认套件
|
|
39
|
+
const defaultCmd = instance.FAIL_TO_PASS.length > 0 ? `npm test -- ${instance.FAIL_TO_PASS.join(' ')}` : 'npm test';
|
|
40
|
+
const cmd = (0, exec_1.sanitizeManagedCommand)(this.opts.testCommand, defaultCmd);
|
|
41
|
+
if (!cmd)
|
|
42
|
+
return false; // 命令未通过受管约束,判为不通过
|
|
43
|
+
const res = await (0, exec_1.runCommand)(cmd, cwd, this.opts.timeoutMs ?? 120000);
|
|
44
|
+
return res.code === 0;
|
|
45
|
+
}
|
|
46
|
+
}
|
|
47
|
+
exports.TestVerifier = TestVerifier;
|
|
48
|
+
//# sourceMappingURL=verifier.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"verifier.js","sourceRoot":"","sources":["../../src/harness/verifier.ts"],"names":[],"mappings":";;;AAAA;;;;;;;;;;GAUG;AACH,2BAAgC;AAChC,+BAA4B;AAC5B,8CAAyE;AASzE,2BAA2B;AAC3B,MAAa,kBAAkB;IAGA;IAFpB,EAAE,GAAG,aAAa,CAAC;IAE5B,YAA6B,OAAe,aAAa;QAA5B,SAAI,GAAJ,IAAI,CAAwB;IAAG,CAAC;IAE7D,KAAK,CAAC,MAAM,CAAC,GAAW,EAAE,SAA0B;QAClD,OAAO,IAAA,eAAU,EAAC,IAAA,WAAI,EAAC,GAAG,EAAE,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC;IAC1C,CAAC;CACF;AARD,gDAQC;AAED,kCAAkC;AAClC,MAAa,YAAY;IAIJ;IAHV,EAAE,GAAG,UAAU,CAAC;IAEzB,YACmB,OAAqD,EAAE;QAAvD,SAAI,GAAJ,IAAI,CAAmD;IACvE,CAAC;IAEJ,KAAK,CAAC,MAAM,CAAC,GAAW,EAAE,QAAyB;QACjD,oDAAoD;QACpD,MAAM,UAAU,GACd,QAAQ,CAAC,YAAY,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,eAAe,QAAQ,CAAC,YAAY,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC,UAAU,CAAC;QACnG,MAAM,GAAG,GAAG,IAAA,6BAAsB,EAAC,IAAI,CAAC,IAAI,CAAC,WAAW,EAAE,UAAU,CAAC,CAAC;QACtE,IAAI,CAAC,GAAG;YAAE,OAAO,KAAK,CAAC,CAAC,kBAAkB;QAC1C,MAAM,GAAG,GAAG,MAAM,IAAA,iBAAU,EAAC,GAAG,EAAE,GAAG,EAAE,IAAI,CAAC,IAAI,CAAC,SAAS,IAAI,MAAM,CAAC,CAAC;QACtE,OAAO,GAAG,CAAC,IAAI,KAAK,CAAC,CAAC;IACxB,CAAC;CACF;AAhBD,oCAgBC"}
|
package/dist/hello.js
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.hello = hello;
|
|
4
|
+
/**
|
|
5
|
+
* Hello world 示例
|
|
6
|
+
*/
|
|
7
|
+
function hello(name) {
|
|
8
|
+
return `Hello, ${name}!`;
|
|
9
|
+
}
|
|
10
|
+
if (require.main === module) {
|
|
11
|
+
const name = process.argv[2] ?? 'World';
|
|
12
|
+
console.log(hello(name));
|
|
13
|
+
}
|
|
14
|
+
//# sourceMappingURL=hello.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"hello.js","sourceRoot":"","sources":["../src/hello.ts"],"names":[],"mappings":";;AAGA,sBAEC;AALD;;GAEG;AACH,SAAgB,KAAK,CAAC,IAAY;IAChC,OAAO,UAAU,IAAI,GAAG,CAAC;AAC3B,CAAC;AAED,IAAI,OAAO,CAAC,IAAI,KAAK,MAAM,EAAE,CAAC;IAC5B,MAAM,IAAI,GAAG,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,OAAO,CAAC;IACxC,OAAO,CAAC,GAAG,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC;AAC3B,CAAC"}
|