feihong-code 7.6.0 → 8.0.1
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/ISSUE_TEMPLATE/bug_report.md +36 -22
- package/.github/ISSUE_TEMPLATE/config.yml +10 -13
- package/.github/ISSUE_TEMPLATE/feature_request.md +38 -13
- package/.github/PULL_REQUEST_TEMPLATE.md +65 -30
- package/.github/workflows/ci.yml +2 -2
- package/.github/workflows/release.yml +5 -5
- package/.github/workflows/swebench-real.yml +2 -2
- package/CONTRIBUTING.md +52 -3
- package/README.md +656 -636
- package/agent-team/.env.example +13 -0
- package/agent-team/README.md +42 -0
- package/agent-team/agents/__init__.py +0 -0
- package/agent-team/agents/l1_acquisition.py +207 -0
- package/agent-team/agents/l2_content.py +237 -0
- package/agent-team/agents/l3_conversion.py +148 -0
- package/agent-team/agents/l4_analysis.py +118 -0
- package/agent-team/agents/l5_management.py +52 -0
- package/agent-team/agents/registry.py +60 -0
- package/agent-team/config.py +39 -0
- package/agent-team/core/__init__.py +0 -0
- package/agent-team/core/agent.py +73 -0
- package/agent-team/core/models.py +347 -0
- package/agent-team/core/scheduler.py +150 -0
- package/agent-team/core/task.py +121 -0
- package/agent-team/feihongzhi.db +0 -0
- package/agent-team/llm/__init__.py +12 -0
- package/agent-team/llm/base.py +22 -0
- package/agent-team/llm/doubao_llm.py +52 -0
- package/agent-team/llm/mock_llm.py +42 -0
- package/agent-team/main.py +235 -0
- package/agent-team/requirements.txt +2 -0
- package/agent-team/skill_packager.py +248 -0
- package/agent-team/skills/01_AI/350/216/267/345/256/242/347/272/277/347/264/242/351/233/267/350/276/276/SKILL.md +40 -0
- package/agent-team/skills/01_AI/350/216/267/345/256/242/347/272/277/347/264/242/351/233/267/350/276/276/run.py +111 -0
- package/agent-team/skills/02_AI/351/224/200/345/224/256/350/267/237/350/277/233/351/224/200/345/206/240/SKILL.md +38 -0
- package/agent-team/skills/02_AI/351/224/200/345/224/256/350/267/237/350/277/233/351/224/200/345/206/240/run.py +108 -0
- package/agent-team/skills/03_/350/257/204/350/256/272/345/214/272/345/225/206/346/234/272/350/257/206/345/210/253/345/256/230/SKILL.md +38 -0
- package/agent-team/skills/03_/350/257/204/350/256/272/345/214/272/345/225/206/346/234/272/350/257/206/345/210/253/345/256/230/run.py +106 -0
- package/agent-team/skills/04_/345/256/242/346/210/267/350/201/212/345/244/251/346/210/220/344/272/244/345/210/206/346/236/220/345/256/230/SKILL.md +37 -0
- package/agent-team/skills/04_/345/256/242/346/210/267/350/201/212/345/244/251/346/210/220/344/272/244/345/210/206/346/236/220/345/256/230/run.py +107 -0
- package/agent-team/skills/05_/345/220/214/350/241/214/347/210/206/346/254/276/351/233/267/350/276/276/SKILL.md +38 -0
- package/agent-team/skills/05_/345/220/214/350/241/214/347/210/206/346/254/276/351/233/267/350/276/276/run.py +99 -0
- package/agent-team/skills/06_/346/257/217/346/227/245/347/262/276/345/207/206/350/216/267/345/256/242/351/200/211/351/242/230/345/256/230/SKILL.md +39 -0
- package/agent-team/skills/06_/346/257/217/346/227/245/347/262/276/345/207/206/350/216/267/345/256/242/351/200/211/351/242/230/345/256/230/run.py +103 -0
- package/agent-team/skills/07_AI/345/244/226/350/264/270/345/274/200/345/217/221/345/256/242/346/210/267/345/221/230/345/267/245/SKILL.md +39 -0
- package/agent-team/skills/07_AI/345/244/226/350/264/270/345/274/200/345/217/221/345/256/242/346/210/267/345/221/230/345/267/245/run.py +103 -0
- package/agent-team/skills/08_/345/260/217/347/272/242/344/271/246/347/262/276/345/207/206/350/216/267/345/256/242/345/221/230/345/267/245/SKILL.md +40 -0
- package/agent-team/skills/08_/345/260/217/347/272/242/344/271/246/347/262/276/345/207/206/350/216/267/345/256/242/345/221/230/345/267/245/run.py +97 -0
- package/agent-team/skills/09_/347/247/201/345/237/237/345/256/242/346/210/267/345/210/206/345/261/202/350/267/237/350/277/233/345/256/230/SKILL.md +38 -0
- package/agent-team/skills/09_/347/247/201/345/237/237/345/256/242/346/210/267/345/210/206/345/261/202/350/267/237/350/277/233/345/256/230/run.py +106 -0
- package/agent-team/skills/10_GEO/345/225/206/346/234/272/350/257/212/346/226/255/345/256/230/SKILL.md +38 -0
- package/agent-team/skills/10_GEO/345/225/206/346/234/272/350/257/212/346/226/255/345/256/230/run.py +108 -0
- package/agent-team/skills/11_/344/270/200/344/272/272/345/205/254/345/217/270AI/345/221/230/345/267/245/350/260/203/345/272/246/345/256/230/SKILL.md +37 -0
- package/agent-team/skills/11_/344/270/200/344/272/272/345/205/254/345/217/270AI/345/221/230/345/267/245/350/260/203/345/272/246/345/256/230/run.py +111 -0
- package/agent-team/skills/12_/350/200/201/346/235/277/347/237/255/350/247/206/351/242/221/350/216/267/345/256/242/350/204/232/346/234/254/345/256/230/SKILL.md +39 -0
- package/agent-team/skills/12_/350/200/201/346/235/277/347/237/255/350/247/206/351/242/221/350/216/267/345/256/242/350/204/232/346/234/254/345/256/230/run.py +104 -0
- package/agent-team/skills/13_/345/215/232/344/270/273/350/264/246/345/217/267/346/267/261/345/272/246/346/213/206/350/247/243/345/256/230/SKILL.md +37 -0
- package/agent-team/skills/13_/345/215/232/344/270/273/350/264/246/345/217/267/346/267/261/345/272/246/346/213/206/350/247/243/345/256/230/run.py +105 -0
- package/agent-team/skills/14_AI/347/237/255/350/247/206/351/242/221/350/207/252/345/212/250/345/211/252/350/276/221/345/270/210/SKILL.md +39 -0
- package/agent-team/skills/14_AI/347/237/255/350/247/206/351/242/221/350/207/252/345/212/250/345/211/252/350/276/221/345/270/210/run.py +108 -0
- package/agent-team/skills/15_/345/205/254/344/274/227/345/217/267/350/207/252/345/212/250/350/277/220/350/220/245/345/217/221/345/270/203/345/256/230/SKILL.md +38 -0
- package/agent-team/skills/15_/345/205/254/344/274/227/345/217/267/350/207/252/345/212/250/350/277/220/350/220/245/345/217/221/345/270/203/345/256/230/run.py +105 -0
- package/agent-team/skills/16_/350/241/214/344/270/232/346/267/261/345/272/246/345/210/206/346/236/220/345/270/210/SKILL.md +38 -0
- package/agent-team/skills/16_/350/241/214/344/270/232/346/267/261/345/272/246/345/210/206/346/236/220/345/270/210/run.py +108 -0
- package/agent-team/skills/17_/346/212/225/350/265/204/347/240/224/347/251/266/345/222/250/350/257/242/345/256/230/SKILL.md +38 -0
- package/agent-team/skills/17_/346/212/225/350/265/204/347/240/224/347/251/266/345/222/250/350/257/242/345/256/230/run.py +106 -0
- package/agent-team/tests/__init__.py +0 -0
- package/agent-team/tests/test_core.py +156 -0
- package/dist/agent/agent-session.js +103 -0
- package/dist/agent/agent-session.js.map +1 -0
- package/dist/agent/orchestrator.js +8 -0
- package/dist/agent/orchestrator.js.map +1 -1
- package/dist/agent/steer.js +52 -0
- package/dist/agent/steer.js.map +1 -0
- package/dist/cli/run.js +2 -0
- package/dist/cli/run.js.map +1 -1
- package/dist/cli/version.js +1 -1
- package/dist/enterprise/policy.js +40 -17
- package/dist/enterprise/policy.js.map +1 -1
- package/dist/memory/honcho-store.js +231 -0
- package/dist/memory/honcho-store.js.map +1 -0
- package/dist/shared/log-rotator.js +238 -0
- package/dist/shared/log-rotator.js.map +1 -0
- package/dist/shared/sqlite-store.js +512 -0
- package/dist/shared/sqlite-store.js.map +1 -0
- package/dist/tools/agents/agents-bridge.js +144 -0
- package/dist/tools/agents/agents-bridge.js.map +1 -0
- package/dist/tools/desktop/desktop-guard.js +81 -0
- package/dist/tools/desktop/desktop-guard.js.map +1 -0
- package/dist/tools/desktop/desktop.tool.js +201 -0
- package/dist/tools/desktop/desktop.tool.js.map +1 -0
- package/dist/tools/desktop/feihong-win.js +163 -0
- package/dist/tools/desktop/feihong-win.js.map +1 -0
- package/dist/tools/desktop/index.js +37 -0
- package/dist/tools/desktop/index.js.map +1 -0
- package/dist/tools/docker-sandbox.js +212 -0
- package/dist/tools/docker-sandbox.js.map +1 -0
- package/dist/tools/index.js +2 -0
- package/dist/tools/index.js.map +1 -1
- package/dist/tools/mcp/index.js +10 -2
- package/dist/tools/mcp/index.js.map +1 -1
- package/dist/tools/mcp/mcp-client.js +4 -0
- package/dist/tools/mcp/mcp-client.js.map +1 -1
- package/dist/tools/sandbox.js +28 -0
- package/dist/tools/sandbox.js.map +1 -1
- package/dist/web/public/favicon.svg +11 -0
- package/dist/web/public/index.html +4 -1
- package/dist/web/public/public/css/style.css +1756 -0
- package/dist/web/public/public/favicon.svg +11 -0
- package/dist/web/public/public/index.html +1246 -0
- package/dist/web/public/public/js/api.js +558 -0
- package/dist/web/public/public/js/app.js +2233 -0
- package/dist/web/public/public/js/mock-api.js +259 -0
- package/dist/web/public/public/js/monaco-editor.js +363 -0
- package/dist/web/public/public/js/ui.js +2217 -0
- package/dist/web/public/public/js/utils.js +214 -0
- package/dist/web/public/public/manifest.json +59 -0
- package/dist/web/public/public/robots.txt +14 -0
- package/dist/web/public/public/service-worker.js +149 -0
- package/dist/web/public/robots.txt +14 -0
- package/dist/web/server.js +70 -0
- package/dist/web/server.js.map +1 -1
- package/docs/AI/345/221/230/345/267/245/345/206/233/345/233/242-/346/225/264/345/220/210/350/257/264/346/230/216.md +68 -0
- package/docs/API.md +827 -0
- package/docs/App/346/212/200/346/234/257/350/257/264/346/230/216/344/271/246.md +3 -3
- package/docs/P1.1-P3-P5/345/267/245/347/250/213/345/256/236/346/226/275/346/200/273/347/273/223.md +328 -0
- package/docs/P5-Pi/345/257/271/347/205/247/350/257/204/346/265/213/346/212/245/345/221/212-mock/346/250/241/345/274/217.json +125 -0
- package/docs/P5-Pi/345/257/271/347/205/247/350/257/204/346/265/213/346/212/245/345/221/212-/347/234/237/345/256/236/346/250/241/345/236/213.json +138 -0
- package/docs/P5-pi-compare-real-report.json +138 -0
- package/docs/P5-pi-compare-report.json +125 -0
- package/docs/PROJECT-ANALYSIS-AND-TEST.md +721 -0
- package/docs/Pi-Agent/346/216/245/345/205/245/351/243/236/350/231/271Code-MVP/351/233/206/346/210/220/344/272/244/344/273/230/347/211/251.md +355 -0
- package/docs/Pi-Agent/346/216/245/345/205/245/351/243/236/350/231/271Code/344/273/267/345/200/274/345/210/206/346/236/220-/345/274/200/345/217/221/346/226/207/346/241/243.md +291 -0
- package/docs/Pi-Agent/346/216/245/345/205/245/351/243/236/350/231/271Code/344/273/267/345/200/274/351/207/215/344/274/260-/345/217/257/350/247/206/345/214/226.html +84 -0
- package/docs/SWE-bench-300-agnes-2.5-flash-/346/234/200/347/273/210/346/212/245/345/221/212.json +2012 -0
- package/docs/SWE-bench-300-agnes-2.5-flash-/350/267/221/345/210/206/346/212/245/345/221/212.md +74 -0
- package/docs/SWE-bench-Lite-DeepSeek-V4-Flash-/350/267/221/345/210/206/346/212/245/345/221/212.md +102 -0
- package/docs/TECHNICAL-SPECIFICATION.md +812 -0
- package/docs/TECHNICAL-SPECIFICATION.pdf +0 -0
- package/docs/UPGRADE-v7.9.0.md +291 -0
- package/docs/UPGRADE-v8.0.0.md +237 -0
- package/docs/Windows/346/211/213/350/204/232-POC/346/274/224/347/244/272/350/204/232/346/234/254.md +131 -0
- package/docs/Windows/346/211/213/350/204/232/346/216/245/345/205/245/346/214/207/345/215/227.md +181 -0
- package/docs//344/275/277/347/224/250/350/257/264/346/230/216/344/271/246-v8.0.0.md +922 -0
- package/docs//344/275/277/347/224/250/350/257/264/346/230/216/344/271/246-v8.0.1.md +146 -0
- package/package.json +7 -6
|
@@ -0,0 +1,138 @@
|
|
|
1
|
+
{
|
|
2
|
+
"date": "2026-09-05T13:01:28.295Z",
|
|
3
|
+
"mode": "real",
|
|
4
|
+
"model": "agnes-2.5-flash",
|
|
5
|
+
"fhcodeVersion": "8.0.0",
|
|
6
|
+
"piVersion": "0.85.1",
|
|
7
|
+
"tasks": [
|
|
8
|
+
{
|
|
9
|
+
"id": "R1-create-file",
|
|
10
|
+
"description": "真实模型:创建文件任务"
|
|
11
|
+
},
|
|
12
|
+
{
|
|
13
|
+
"id": "R2-read-summarize",
|
|
14
|
+
"description": "真实模型:读取并总结"
|
|
15
|
+
},
|
|
16
|
+
{
|
|
17
|
+
"id": "R3-steer-real",
|
|
18
|
+
"description": "真实模型:中途纠偏"
|
|
19
|
+
}
|
|
20
|
+
],
|
|
21
|
+
"metrics": [
|
|
22
|
+
{
|
|
23
|
+
"taskId": "R1-create-file",
|
|
24
|
+
"agent": "fhcode",
|
|
25
|
+
"iterations": 1,
|
|
26
|
+
"toolCalls": 1,
|
|
27
|
+
"success": true,
|
|
28
|
+
"durationMs": 5139,
|
|
29
|
+
"steerInjected": false,
|
|
30
|
+
"steerObserved": false,
|
|
31
|
+
"finalAnswer": "文件已成功创建。我已在当前目录下创建了 hello.ts 文件,内容为 console.log(\"Hello, 飞虹Code!\");。这个 TypeScript 文件可以运行并输出\"Hello, 飞虹Code!\"到控制台。",
|
|
32
|
+
"verifyReason": "hello.ts 已创建且内容正确"
|
|
33
|
+
},
|
|
34
|
+
{
|
|
35
|
+
"taskId": "R1-create-file",
|
|
36
|
+
"agent": "pi",
|
|
37
|
+
"iterations": 2,
|
|
38
|
+
"toolCalls": 1,
|
|
39
|
+
"success": true,
|
|
40
|
+
"durationMs": 1638,
|
|
41
|
+
"steerInjected": false,
|
|
42
|
+
"steerObserved": false,
|
|
43
|
+
"finalAnswer": "[{\"type\":\"text\",\"text\":\"已创建 `hello.ts` 文件,内容为:\\n\\n```typescript\\nconsole.log(\\\"Hello, 飞虹Code!\\\");\\n```\\n\\n使用 `npx ts-node hello.ts` 或 `deno run hello.",
|
|
44
|
+
"verifyReason": "hello.ts 已创建且内容正确"
|
|
45
|
+
},
|
|
46
|
+
{
|
|
47
|
+
"taskId": "R2-read-summarize",
|
|
48
|
+
"agent": "fhcode",
|
|
49
|
+
"iterations": 1,
|
|
50
|
+
"toolCalls": 1,
|
|
51
|
+
"success": true,
|
|
52
|
+
"durationMs": 12593,
|
|
53
|
+
"steerInjected": false,
|
|
54
|
+
"steerObserved": false,
|
|
55
|
+
"finalAnswer": "项目名称是 feihong-code,版本号为 8.0.0。",
|
|
56
|
+
"verifyReason": "任务已执行"
|
|
57
|
+
},
|
|
58
|
+
{
|
|
59
|
+
"taskId": "R2-read-summarize",
|
|
60
|
+
"agent": "pi",
|
|
61
|
+
"iterations": 2,
|
|
62
|
+
"toolCalls": 1,
|
|
63
|
+
"success": true,
|
|
64
|
+
"durationMs": 29920,
|
|
65
|
+
"steerInjected": false,
|
|
66
|
+
"steerObserved": false,
|
|
67
|
+
"finalAnswer": "[{\"type\":\"text\",\"text\":\"项目名称是 feihong-code,版本号为 8.0.0。\"}]",
|
|
68
|
+
"verifyReason": "任务已执行"
|
|
69
|
+
},
|
|
70
|
+
{
|
|
71
|
+
"taskId": "R3-steer-real",
|
|
72
|
+
"agent": "fhcode",
|
|
73
|
+
"iterations": 2,
|
|
74
|
+
"toolCalls": 2,
|
|
75
|
+
"success": true,
|
|
76
|
+
"durationMs": 30179,
|
|
77
|
+
"steerInjected": true,
|
|
78
|
+
"steerObserved": true,
|
|
79
|
+
"finalAnswer": "已在当前目录创建 result.md 文件,内容为\"新内容\"。原有的 test.txt 文件也保持不变。",
|
|
80
|
+
"verifyReason": "steer 生效:result.md 已创建"
|
|
81
|
+
},
|
|
82
|
+
{
|
|
83
|
+
"taskId": "R3-steer-real",
|
|
84
|
+
"agent": "pi",
|
|
85
|
+
"iterations": 3,
|
|
86
|
+
"toolCalls": 2,
|
|
87
|
+
"success": true,
|
|
88
|
+
"durationMs": 2192,
|
|
89
|
+
"steerInjected": true,
|
|
90
|
+
"steerObserved": true,
|
|
91
|
+
"finalAnswer": "[{\"type\":\"text\",\"text\":\"已创建 `result.md` 文件,内容为\\\"新内容\\\"。\"}]",
|
|
92
|
+
"verifyReason": "steer 生效:result.md 已创建"
|
|
93
|
+
}
|
|
94
|
+
],
|
|
95
|
+
"comparison": [
|
|
96
|
+
{
|
|
97
|
+
"task": "R1-create-file",
|
|
98
|
+
"fhcode 迭代": 1,
|
|
99
|
+
"pi 迭代": 2,
|
|
100
|
+
"fhcode 工具": 1,
|
|
101
|
+
"pi 工具": 1,
|
|
102
|
+
"fhcode 耗时ms": 5139,
|
|
103
|
+
"pi 耗时ms": 1638,
|
|
104
|
+
"fhcode 成功": true,
|
|
105
|
+
"pi 成功": true,
|
|
106
|
+
"fhcode 验证": "hello.ts 已创建且内容正确",
|
|
107
|
+
"pi 验证": "hello.ts 已创建且内容正确"
|
|
108
|
+
},
|
|
109
|
+
{
|
|
110
|
+
"task": "R2-read-summarize",
|
|
111
|
+
"fhcode 迭代": 1,
|
|
112
|
+
"pi 迭代": 2,
|
|
113
|
+
"fhcode 工具": 1,
|
|
114
|
+
"pi 工具": 1,
|
|
115
|
+
"fhcode 耗时ms": 12593,
|
|
116
|
+
"pi 耗时ms": 29920,
|
|
117
|
+
"fhcode 成功": true,
|
|
118
|
+
"pi 成功": true,
|
|
119
|
+
"fhcode 验证": "任务已执行",
|
|
120
|
+
"pi 验证": "任务已执行"
|
|
121
|
+
},
|
|
122
|
+
{
|
|
123
|
+
"task": "R3-steer-real",
|
|
124
|
+
"fhcode 迭代": 2,
|
|
125
|
+
"pi 迭代": 3,
|
|
126
|
+
"fhcode 工具": 2,
|
|
127
|
+
"pi 工具": 2,
|
|
128
|
+
"fhcode 耗时ms": 30179,
|
|
129
|
+
"pi 耗时ms": 2192,
|
|
130
|
+
"fhcode 成功": true,
|
|
131
|
+
"pi 成功": true,
|
|
132
|
+
"fhcode steer": true,
|
|
133
|
+
"pi steer": true,
|
|
134
|
+
"fhcode 验证": "steer 生效:result.md 已创建",
|
|
135
|
+
"pi 验证": "steer 生效:result.md 已创建"
|
|
136
|
+
}
|
|
137
|
+
]
|
|
138
|
+
}
|
|
@@ -0,0 +1,125 @@
|
|
|
1
|
+
{
|
|
2
|
+
"date": "2026-09-05T12:25:15.717Z",
|
|
3
|
+
"mode": "mock",
|
|
4
|
+
"fhcodeVersion": "8.0.0",
|
|
5
|
+
"piVersion": "0.85.1",
|
|
6
|
+
"tasks": [
|
|
7
|
+
{
|
|
8
|
+
"id": "T1-single-tool",
|
|
9
|
+
"description": "单工具调用任务:读文件后总结"
|
|
10
|
+
},
|
|
11
|
+
{
|
|
12
|
+
"id": "T2-multi-tool",
|
|
13
|
+
"description": "多工具链式任务:读文件 → 写文件 → 总结"
|
|
14
|
+
},
|
|
15
|
+
{
|
|
16
|
+
"id": "T3-steer-correction",
|
|
17
|
+
"description": "中途纠偏任务:执行到一半时用户改方向"
|
|
18
|
+
}
|
|
19
|
+
],
|
|
20
|
+
"metrics": [
|
|
21
|
+
{
|
|
22
|
+
"taskId": "T1-single-tool",
|
|
23
|
+
"agent": "fhcode",
|
|
24
|
+
"iterations": 1,
|
|
25
|
+
"toolCalls": 1,
|
|
26
|
+
"success": true,
|
|
27
|
+
"durationMs": 35,
|
|
28
|
+
"steerInjected": false,
|
|
29
|
+
"steerObserved": false,
|
|
30
|
+
"finalAnswer": "已读取文件,包含 3 个导出函数。"
|
|
31
|
+
},
|
|
32
|
+
{
|
|
33
|
+
"taskId": "T1-single-tool",
|
|
34
|
+
"agent": "pi",
|
|
35
|
+
"iterations": 2,
|
|
36
|
+
"toolCalls": 1,
|
|
37
|
+
"success": true,
|
|
38
|
+
"durationMs": 7,
|
|
39
|
+
"steerInjected": false,
|
|
40
|
+
"steerObserved": false,
|
|
41
|
+
"finalAnswer": "[{\"type\":\"text\",\"text\":\"已读取文件,包含 3 个导出函数。\"}]"
|
|
42
|
+
},
|
|
43
|
+
{
|
|
44
|
+
"taskId": "T2-multi-tool",
|
|
45
|
+
"agent": "fhcode",
|
|
46
|
+
"iterations": 2,
|
|
47
|
+
"toolCalls": 2,
|
|
48
|
+
"success": true,
|
|
49
|
+
"durationMs": 25,
|
|
50
|
+
"steerInjected": false,
|
|
51
|
+
"steerObserved": false,
|
|
52
|
+
"finalAnswer": "已完成:读取 a.ts,创建 b.ts。"
|
|
53
|
+
},
|
|
54
|
+
{
|
|
55
|
+
"taskId": "T2-multi-tool",
|
|
56
|
+
"agent": "pi",
|
|
57
|
+
"iterations": 3,
|
|
58
|
+
"toolCalls": 2,
|
|
59
|
+
"success": true,
|
|
60
|
+
"durationMs": 2,
|
|
61
|
+
"steerInjected": false,
|
|
62
|
+
"steerObserved": false,
|
|
63
|
+
"finalAnswer": "[{\"type\":\"text\",\"text\":\"已完成:读取 a.ts,创建 b.ts。\"}]"
|
|
64
|
+
},
|
|
65
|
+
{
|
|
66
|
+
"taskId": "T3-steer-correction",
|
|
67
|
+
"agent": "fhcode",
|
|
68
|
+
"iterations": 1,
|
|
69
|
+
"toolCalls": 1,
|
|
70
|
+
"success": true,
|
|
71
|
+
"durationMs": 25,
|
|
72
|
+
"steerInjected": true,
|
|
73
|
+
"steerObserved": true,
|
|
74
|
+
"finalAnswer": "已按新指令完成:改为读取 new.ts 并总结。"
|
|
75
|
+
},
|
|
76
|
+
{
|
|
77
|
+
"taskId": "T3-steer-correction",
|
|
78
|
+
"agent": "pi",
|
|
79
|
+
"iterations": 2,
|
|
80
|
+
"toolCalls": 1,
|
|
81
|
+
"success": true,
|
|
82
|
+
"durationMs": 2,
|
|
83
|
+
"steerInjected": true,
|
|
84
|
+
"steerObserved": true,
|
|
85
|
+
"finalAnswer": "[{\"type\":\"text\",\"text\":\"已按新指令完成:改为读取 new.ts 并总结。\"}]"
|
|
86
|
+
}
|
|
87
|
+
],
|
|
88
|
+
"comparison": [
|
|
89
|
+
{
|
|
90
|
+
"task": "T1-single-tool",
|
|
91
|
+
"fhcode 迭代": 1,
|
|
92
|
+
"pi 迭代": 2,
|
|
93
|
+
"fhcode 工具调用": 1,
|
|
94
|
+
"pi 工具调用": 1,
|
|
95
|
+
"fhcode 耗时(ms)": 35,
|
|
96
|
+
"pi 耗时(ms)": 7,
|
|
97
|
+
"fhcode 成功": true,
|
|
98
|
+
"pi 成功": true
|
|
99
|
+
},
|
|
100
|
+
{
|
|
101
|
+
"task": "T2-multi-tool",
|
|
102
|
+
"fhcode 迭代": 2,
|
|
103
|
+
"pi 迭代": 3,
|
|
104
|
+
"fhcode 工具调用": 2,
|
|
105
|
+
"pi 工具调用": 2,
|
|
106
|
+
"fhcode 耗时(ms)": 25,
|
|
107
|
+
"pi 耗时(ms)": 2,
|
|
108
|
+
"fhcode 成功": true,
|
|
109
|
+
"pi 成功": true
|
|
110
|
+
},
|
|
111
|
+
{
|
|
112
|
+
"task": "T3-steer-correction",
|
|
113
|
+
"fhcode 迭代": 1,
|
|
114
|
+
"pi 迭代": 2,
|
|
115
|
+
"fhcode 工具调用": 1,
|
|
116
|
+
"pi 工具调用": 1,
|
|
117
|
+
"fhcode 耗时(ms)": 25,
|
|
118
|
+
"pi 耗时(ms)": 2,
|
|
119
|
+
"fhcode 成功": true,
|
|
120
|
+
"pi 成功": true,
|
|
121
|
+
"fhcode steer生效": true,
|
|
122
|
+
"pi steer生效": true
|
|
123
|
+
}
|
|
124
|
+
]
|
|
125
|
+
}
|