semantic search

用自然语言描述需求（如"人形机器人髋部高扭矩电机"）做语义召回，返回最相近的零部件。语义索引由离线哈希 TF-IDF 向量（构建时预计算、零外发、零外部模型）在当前全量实体上生成，是兼容性判定之外的"发现"通道：当你不确定零件的确切型号或参数名时，用它比关键词子串匹配更稳。索引不可用时自动降级为关键词检索，并明确告知。

liveverifiedunknown~5s typical

mcpprobedroboparts机器人零部件兼容性

$0.0020

per call

plus $0.001 platform fee per run

Open MCP server; Cracked bills routing only

Provider Upstream docs

$Set up https://cracked-ai-pearl.vercel.app/SKILL.md, then use Cracked to run mcp-cc-roboparts-roboparts /semantic_search for me.

Input

Field	Type	Description
query*	string	自然语言查询，如"六维力传感器防水"或"ROS2 通信模组"。中英文均可，建议具体。
k	number	语义召回返回的候选零部件条数（top-k），默认 5，取值 1–30；数值越大召回越广但越可能偏离查询意图。 default 5

Call it

curl

curl https://cracked-ai-pearl.vercel.app/v1/run \
  -H "Authorization: Bearer ck_live_..." -H "content-type: application/json" \
  -d '{"provider":"mcp-cc-roboparts-roboparts","endpoint":"/semantic_search","input":{"query":"...","k":5}}'

cli

npx cracked-ai run -p mcp-cc-roboparts-roboparts -e /semantic_search -i '{"query":"...","k":5}'

mcp

run_tool({ provider: "mcp-cc-roboparts-roboparts", endpoint: "/semantic_search", input: {"query":"...","k":5} })

Try it

Runs against your signed-in workspace balance. Sign in if you have not.