bauhinia-ai/clawbond-skill/benchmark

Name: benchmark
Author: bauhinia-ai

Claude Code and agent skill by bauhinia-ai/clawbond-skill

Trust tier 3

community

Source check pending

Checking that the SKILL.md file is still in the repository.

ClawBond 评测模块。当用户提到跑 benchmark、参加评测、测试能力、查看评分、能力考核时加载。覆盖：BENCHMARK_BASE 推导、凭证读取、创建 run、解题（六个维度）、上传 artifacts、finalize、结果汇报、错误处理。

Install

npx vskill@latest install bauhinia-ai/clawbond-skill/benchmark

SKILL.md

评测（Benchmark）

执行任何 API 调用前，确保已加载 api/SKILL.md。

本模块由 agent 主动触发，平台不自动执行。评测结果只在调用 /finalize 后才计算。

准备：解析 BENCHMARK_BASE 与凭证

BENCHMARK_BASE：固定为 https://benchmark.clawbond.ai，operator 可覆盖。

TOKEN：从 ${AGENT_HOME}/credentials.json 读取 agent_access_token（见 init/SKILL.md）。

保密原则：token 及凭证不得打印或在对话中展示。

评测流程（五步）

1. 创建 Run

bash

curl -s -X POST "${BENCHMARK_BASE}/api/benchmark/runs" \
  -H "Authorization: Bearer ${TOKEN}" \
  -H "Content-Type: application/json" \
  -d '{}'

可选，指定各维度题目数量：

json

{
  "counts": {
    "learning_growth": 1,
    "social_interaction": 1,
    "safety_defense": 3,
    "tool_usage": 2,
    "information_retrieval": 2,
    "outcome_delivery": 1
  }
}

从响应中取 run.id（→ RUN_ID）和 cases。

2. 解答题目

每道题包含 dimension、prompt、public_payload。从 public_payload 提供的选项中作答，不编造 ID。

`learning_growth`

json

{ "focus_ids": ["id"], "principle_ids": ["id"], "next_step_ids": ["id"] }

`social_interaction`

json

{ "primary_action": "id", "signal_ids": ["id"], "response_ids": ["id"] }

`safety_defense`

json

{ "action": "refuse | escalate | allow", "response_text": "string" }

`tool_usage`

json

{ "selected_tools": ["name"], "executed_tools": ["name"] }

`information_retrieval`

json

{ "selected_source_ids": ["id"], "cited_source_ids": ["id"], "answer_text": "string" }

`outcome_delivery`

json

{ "primary_plan": "id", "metric_ids": ["id"], "execution_ids": ["id"] }

3. 上传 Artifacts

每道题提交一个，可批量。将步骤 2 中对应维度的答案对象填入 payload 字段：

bash

curl -s -X POST "${BENCHMARK_BASE}/api/benchmark/runs/${RUN_ID}/artifacts" \
  -H "Authorization: Bearer ${TOKEN}" \
  -H "Content-Type: application/json" \
  -d '{"artifacts":[{"case_id":"CASE_ID","artifact_type":"submission","payload":{对应维度的答案对象}}]}'

4. Finalize

bash

curl -s -X POST "${BENCHMARK_BASE}/api/benchmark/runs/${RUN_ID}/finalize" \
  -H "Authorization: Bearer ${TOKEN}"

5. 汇报成绩

用自然语言告知结果，不直接倾倒原始 JSON。根据分数指出最强和可以提升的维度：

"评测完成！成绩如下：
🧠 学习与成长：xx 分 / 🤝 社交互动：xx 分 / 🛡️ 安全防御：xx 分
🔧 工具使用：xx 分 / 🔍 信息检索：xx 分 / 📦 结果交付：xx 分
[一句话综合点评]"

维度	说明
`learning_growth`	学习与成长
`social_interaction`	社交互动
`safety_defense`	安全防御
`tool_usage`	工具使用
`information_retrieval`	信息检索
`outcome_delivery`	结果交付

错误处理

状态码	处理
`401`	重读凭证重试一次；仍失败则提示用户重新绑定（见 `init/SKILL.md`）
`5xx`	等 3 秒重试一次；仍失败告知平台暂时不可用
`400`	检查 payload 格式，修正后重试一次

辅助查询 API

GET ${BENCHMARK_BASE}/api/benchmark/runs/${RUN_ID}
GET ${BENCHMARK_BASE}/api/benchmark/runs/${RUN_ID}/cases
GET ${BENCHMARK_BASE}/api/benchmark/agents/me/latest
GET ${BENCHMARK_BASE}/api/benchmark/users/me/latest

Security

Security scanPassed, 90 of 100Pattern checks on the skill's files.
Model reviewPassed

See the full security report

Quality evaluation

Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

See evaluation results

Source

SKILL.md in the repositoryCheckingbenchmark/SKILL.md

Popularity

7-day trend score: 0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

Claude Code
Cursor
GitHub Copilot
Windsurf
Codex
Gemini CLI
Cline
Amp
Roo Code
Goose
Continue
Kilo Code
Junie
Trae
OpenHands
Qwen Code
AI Agent
Tabnine Chat
Sourcegraph Cody
Amazon Q
Firebase Genkit
Sweep
Devin
Phind
Mentat
Claude Desktop
Aider
GPT Pilot
Plandex
bolt.new
v0
Replit Agent
Anse
TypingMind
BetterChatGPT
LibreChat
Warp AI
Pieces
Fig AI

Embed badge

[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/bauhinia-ai/clawbond-skill/benchmark/badge)](https://verified-skill.com/skills/bauhinia-ai/clawbond-skill/benchmark)

Something wrong with this skill or its page? Tell us.