ai-agent-book 精选快照(<2MB 代码与文档,来自 github.com/bojieli/ai-agent-book)
Build latest book artifacts / build (push) Canceled after 0s
dependency resolution / resolve (3.11) (push) Canceled after 0s
dependency resolution / resolve (3.13) (push) Canceled after 0s
deploy-pages / build (push) Canceled after 0s
deploy-pages / deploy (push) Canceled after 0s
i18n consistency check / check (push) Canceled after 0s
provider adoption tests / test (chapter2/context-compression) (push) Canceled after 0s
provider adoption tests / test (chapter2/prompt-injection) (push) Canceled after 0s
provider adoption tests / test (chapter2/system-hint) (push) Canceled after 0s
provider adoption tests / test (chapter3/log-sanitization) (push) Canceled after 0s
web-search-agent tests / test (push) Canceled after 0s
web-search-agent tests / agentbook (push) Canceled after 0s

This commit is contained in:
2026-08-20 13:12:50 +00:00
commit b119135836
10275 changed files with 3284984 additions and 0 deletions
@@ -0,0 +1,508 @@
{
"experiment_id": "2-4",
"created_at": "2026-07-30T05:14:32.789700+08:00",
"protocol_sha256": "ec1ec36d5a6ee21b5185295a7a4447f4e0173f1681a9fe471b240dda1e7cf32a",
"protocol_copy": "/Users/boj/book/ai-agent-book/chapter2/prompt-engineering/runs/exp2-4-kimi-k3-20260730-v2/experiment_protocol.json",
"provider": "openai",
"model": "kimi-k3",
"user_model": "kimi-k3",
"objective_scoring": "vendored tau-bench environment reward",
"arms": {
"baseline": {
"artifact": {
"path": "/Users/boj/book/ai-agent-book/chapter2/prompt-engineering/runs/exp2-4-kimi-k3-20260730-v2/tool-calling-kimi-k3-baseline_0730022423.json",
"sha256": "967697b4cb140efa01230c54ce6b177f227c8a617e785071e60c2d4483b39fba"
},
"rewards": [
0.0,
1.0,
0.0,
1.0,
1.0,
1.0,
1.0,
1.0,
1.0,
0.0
],
"success_rate": 70.0,
"total": 10,
"successes": 7,
"failures": 3,
"tasks_completed": 10,
"expected_tasks": 10,
"task_errors": [
{
"task_id": 9,
"error": "litellm.BadRequestError: OpenAIException - Invalid request: the message at position 25 with role 'user' must not be empty"
}
],
"agent_steps": [
14,
10,
30,
18,
15,
14,
11,
14,
16,
null
],
"tool_calls": [
6,
5,
28,
14,
12,
11,
7,
11,
12,
null
],
"tool_errors": [
1,
0,
0,
0,
0,
0,
0,
0,
0,
null
],
"real_api_calls": 187,
"response_ids_present": true,
"usage": {
"prompt_tokens": 1308761,
"completion_tokens": 71951,
"total_tokens": 1380712
},
"observed_litellm_cost_usd": 0,
"all_calls_priced": false,
"native_cost_cny": 33.37032,
"arm_complete": false
},
"tone_trump": {
"artifact": {
"path": "/Users/boj/book/ai-agent-book/chapter2/prompt-engineering/runs/exp2-4-kimi-k3-20260730-v2/tool-calling-kimi-k3-tone_trump_0730025250.json",
"sha256": "9ed3c56f487b13c9ce47adc46404c4d63d0b28aaf595c4ba42154dcf7e12bfff"
},
"rewards": [
1.0,
1.0,
0.0,
1.0,
0.0,
1.0,
1.0,
1.0,
0.0,
0.0
],
"success_rate": 60.0,
"total": 10,
"successes": 6,
"failures": 4,
"tasks_completed": 10,
"expected_tasks": 10,
"task_errors": [
{
"task_id": 9,
"error": "litellm.BadRequestError: OpenAIException - Invalid request: the message at position 27 with role 'user' must not be empty"
}
],
"agent_steps": [
13,
11,
16,
18,
12,
15,
10,
12,
14,
null
],
"tool_calls": [
7,
7,
14,
14,
9,
12,
6,
8,
10,
null
],
"tool_errors": [
0,
0,
0,
0,
0,
0,
0,
0,
0,
null
],
"real_api_calls": 164,
"response_ids_present": true,
"usage": {
"prompt_tokens": 1100695,
"completion_tokens": 78345,
"total_tokens": 1179040
},
"observed_litellm_cost_usd": 0,
"all_calls_priced": false,
"native_cost_cny": 29.848399999999998,
"arm_complete": false
},
"tone_casual": {
"artifact": {
"path": "/Users/boj/book/ai-agent-book/chapter2/prompt-engineering/runs/exp2-4-kimi-k3-20260730-v2/tool-calling-kimi-k3-tone_casual_0730032323.json",
"sha256": "460fe6d9fa11686e4f6b07bf2d1e565c8f12b03cd84f86ec3efc22855ac2aeb5"
},
"rewards": [
1.0,
1.0,
1.0,
1.0,
1.0,
1.0,
1.0,
1.0,
1.0,
0.0
],
"success_rate": 90.0,
"total": 10,
"successes": 9,
"failures": 1,
"tasks_completed": 10,
"expected_tasks": 10,
"task_errors": [
{
"task_id": 9,
"error": "litellm.BadRequestError: OpenAIException - Invalid request: the message at position 7 with role 'user' must not be empty"
}
],
"agent_steps": [
14,
10,
16,
15,
15,
14,
12,
15,
16,
null
],
"tool_calls": [
9,
5,
13,
11,
11,
11,
8,
9,
13,
null
],
"tool_errors": [
0,
0,
0,
0,
0,
0,
0,
0,
0,
null
],
"real_api_calls": 173,
"response_ids_present": true,
"usage": {
"prompt_tokens": 1176456,
"completion_tokens": 74474,
"total_tokens": 1250930
},
"observed_litellm_cost_usd": 0,
"all_calls_priced": false,
"native_cost_cny": 30.97652,
"arm_complete": false
},
"wiki_random": {
"artifact": {
"path": "/Users/boj/book/ai-agent-book/chapter2/prompt-engineering/runs/exp2-4-kimi-k3-20260730-v2/tool-calling-kimi-k3-wiki_random_0730035037.json",
"sha256": "645afb3fe2be8cc5cb816b50e4654211586b02eff934d676bf95cb5a0f0d30ff"
},
"rewards": [
1.0,
1.0,
0.0,
0.0,
0.0,
1.0,
1.0,
1.0,
1.0,
0.0
],
"success_rate": 60.0,
"total": 10,
"successes": 6,
"failures": 4,
"tasks_completed": 10,
"expected_tasks": 10,
"task_errors": [
{
"task_id": 2,
"error": "litellm.BadRequestError: OpenAIException - Invalid request: the message at position 7 with role 'user' must not be empty"
},
{
"task_id": 3,
"error": "litellm.BadRequestError: OpenAIException - Invalid request: the message at position 5 with role 'user' must not be empty"
},
{
"task_id": 9,
"error": "litellm.BadRequestError: OpenAIException - Invalid request: the message at position 31 with role 'user' must not be empty"
}
],
"agent_steps": [
14,
9,
null,
null,
14,
15,
13,
10,
14,
null
],
"tool_calls": [
7,
5,
null,
null,
10,
13,
10,
7,
10,
null
],
"tool_errors": [
0,
0,
null,
null,
0,
0,
0,
0,
0,
null
],
"real_api_calls": 123,
"response_ids_present": true,
"usage": {
"prompt_tokens": 733921,
"completion_tokens": 54842,
"total_tokens": 788763
},
"observed_litellm_cost_usd": 0,
"all_calls_priced": false,
"native_cost_cny": 20.16262,
"arm_complete": false
},
"no_tool_desc": {
"artifact": {
"path": "/Users/boj/book/ai-agent-book/chapter2/prompt-engineering/runs/exp2-4-kimi-k3-20260730-v2/tool-calling-kimi-k3-no_tool_desc_0730042104.json",
"sha256": "4ee2a432ec120556c4c2b45eafee756ff782b2bd44849f8dba7bfa5da84e70f5"
},
"rewards": [
1.0,
1.0,
0.0,
1.0,
1.0,
1.0,
1.0,
1.0,
1.0,
0.0
],
"success_rate": 80.0,
"total": 10,
"successes": 8,
"failures": 2,
"tasks_completed": 10,
"expected_tasks": 10,
"task_errors": [
{
"task_id": 2,
"error": "litellm.BadRequestError: OpenAIException - Invalid request: the message at position 5 with role 'user' must not be empty"
},
{
"task_id": 9,
"error": "litellm.BadRequestError: OpenAIException - Invalid request: the message at position 21 with role 'user' must not be empty"
}
],
"agent_steps": [
14,
9,
null,
15,
12,
15,
11,
13,
13,
null
],
"tool_calls": [
9,
5,
null,
11,
9,
12,
7,
10,
8,
null
],
"tool_errors": [
0,
0,
null,
0,
0,
0,
0,
0,
0,
null
],
"real_api_calls": 141,
"response_ids_present": true,
"usage": {
"prompt_tokens": 788167,
"completion_tokens": 63372,
"total_tokens": 851539
},
"observed_litellm_cost_usd": 0,
"all_calls_priced": false,
"native_cost_cny": 22.10054,
"arm_complete": false
},
"all_ablations": {
"artifact": {
"path": "/Users/boj/book/ai-agent-book/chapter2/prompt-engineering/runs/exp2-4-kimi-k3-20260730-v2/tool-calling-kimi-k3-tone_casual_wiki_random_no_tool_desc_0730044439.json",
"sha256": "981a51a7e60a1bd3459b1e0baba7979b5c002b330899741b082b383861a5fa5a"
},
"rewards": [
1.0,
1.0,
1.0,
0.0,
1.0,
0.0,
1.0,
1.0,
1.0,
0.0
],
"success_rate": 70.0,
"total": 10,
"successes": 7,
"failures": 3,
"tasks_completed": 10,
"expected_tasks": 10,
"task_errors": [
{
"task_id": 3,
"error": "litellm.BadRequestError: OpenAIException - Invalid request: the message at position 7 with role 'user' must not be empty"
},
{
"task_id": 9,
"error": "litellm.BadRequestError: OpenAIException - Invalid request: the message at position 29 with role 'user' must not be empty"
}
],
"agent_steps": [
13,
10,
16,
null,
16,
9,
12,
15,
12,
null
],
"tool_calls": [
7,
5,
13,
null,
12,
8,
6,
10,
8,
null
],
"tool_errors": [
0,
0,
0,
null,
0,
0,
0,
0,
0,
null
],
"real_api_calls": 145,
"response_ids_present": true,
"usage": {
"prompt_tokens": 850906,
"completion_tokens": 68218,
"total_tokens": 919124
},
"observed_litellm_cost_usd": 0,
"all_calls_priced": false,
"native_cost_cny": 23.83992,
"arm_complete": false
}
},
"credential_scan_passed": true,
"credential_findings": [],
"usage_and_cost": {
"total_real_api_calls": 933,
"prompt_tokens": 5958906,
"completion_tokens": 411202,
"total_tokens": 6370108,
"observed_litellm_cost_usd": 0,
"all_calls_priced": false,
"native_cost_cny": 160.29832000000002,
"native_cost_complete": true,
"qualification": "Prompt cache detail is unavailable in these responses, so every prompt token is conservatively priced as uncached."
},
"campaign_complete": false,
"hypothesis_results": {
"historical_percentages_reproduced": false,
"qualification": "Current fixed ten-task campaign only; compare arm metrics directly."
}
}