MCP MeasureMCP Measure

索引已觀測

SwarmIO

AI that does real work for you: cited web research, plans, finance, horoscopes. Self-serve.

我該用這個嗎

品質與安全性

B
說明品質
95%
結構描述完整度
77%
命名品質
67%
汙染風險
80%
權限相符程度
90%
協定合規性
100%

發現項目（14）

  • HIGHTool poisoning patterns detected
  • MEDIUMTool description contains URL to non-standard domain在 runs.launch 中
  • LOWTool 'modes.list' description lacks action verb在 modes.list 中
  • LOWTool 'market.pulse' doesn't follow camelCase/snake_case在 market.pulse 中
  • LOWTool 'route.classify' doesn't follow camelCase/snake_case在 route.classify 中
  • LOWTool 'runs.launch' doesn't follow camelCase/snake_case在 runs.launch 中
  • LOWTool 'runs.get' doesn't follow camelCase/snake_case在 runs.get 中
  • LOWTool 'runs.wait' doesn't follow camelCase/snake_case在 runs.wait 中
  • LOWTool 'runs.report' doesn't follow camelCase/snake_case在 runs.report 中
  • LOWTool 'runs.cancel' doesn't follow camelCase/snake_case在 runs.cancel 中

根據工具定義與協定合規性的自動化分析。

上下文成本

~2,865Token（工具定義）
~2.1 KB典型回應大小
顯著的注意力影響（128k 上下文的 2.24%）

這是每次將伺服器的工具載入模型上下文時所消耗的約略 token 數量。數量越高，可用於其他工作的注意力就越少。

安裝

一鍵安裝

將以下內容加入你的 `claude_desktop_config.json` 檔案：

{
  "mcpServers": {
    "swarmio": {
      "url": "https://ioswarm.io/mcp"
    }
  }
}

遠端端點

https://ioswarm.io/mcpstreamable-http

它能做什麼

工具清單

工具（12）

🟢 唯讀🟡 寫入🔴 刪除⚪ 未知
🟢market.pulse

FREE, no key, no account: live prices for BTC, ETH, SOL, AMZN and NVDA with the session's change, plus today's news — the platform's newest published news digest parsed into stories with source links, and the market wire's latest headlines. Refreshed every minute (prices) and every ten minutes (news). Call this first to see what the swarm's data looks like before spending anything.

輸入結構描述

{
  "type": "object",
  "properties": {}
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "as_of": {
      "description": "RFC 3339 time this document was rendered.",
      "type": "string"
    },
    "attribution": {
      "type": "string"
    },
    "free": {
      "description": "Always true — this call is never charged.",
      "type": "boolean"
    },
    "news": {
      "additionalProperties": true,
      "properties": {
        "as_of": {
          "type": [
            "string",
            "null"
          ]
        },
        "digest": {
          "description": "The newest published news digest, parsed: {title, as_of, report_url, stories:[{headline, link, source, excerpt}]}. Null when no digest ran in the last 36 hours.",
          "type": [
            "object",
            "null"
          ]
        },
        "headlines": {
          "description": "Latest market-wire headlines: {headline, link, publisher, published, summary}.",
          "items": {
            "additionalProperties": true,
            "type": "object"
          },
          "type": "array"
        }
      },
      "type": "object"
    },
    "next": {
      "additionalProperties": true,
      "description": "How to get the same data over MCP and how to commission a real report.",
      "type": "object"
    },
    "prices": {
      "description": "One entry per symbol that priced this cycle, in list order.",
      "items": {
        "additionalProperties": true,
        "properties": {
          "as_of": {
            "type": "string"
          },
          "change_pct": {
            "description": "Session change in percent; null where the source has no session (crypto spot).",
            "type": [
              "number",
              "null"
            ]
          },
          "currency": {
            "type": "string"
          },
          "price": {
            "type": "number"
          },
          "source": {
            "description": "pyth | yahoo.",
            "type": "string"
          },
          "symbol": {
            "type": "string"
          }
        },
        "required": [
          "symbol",
          "price",
          "currency",
          "source",
          "as_of"
        ],
        "type": "object"
      },
      "type": "array"
    },
    "prices_as_of": {
      "type": [
        "string",
        "null"
      ]
    },
    "symbols": {
      "description": "The fixed symbol list served.",
      "items": {
        "type": "string"
      },
      "type": "array"
    }
  },
  "required": [
    "as_of",
    "prices",
    "news"
  ],
  "additionalProperties": true
}
🟢search(query)

Search the swarm's published research: the featured shelf, the operator's daily news / markets / AI digests, and — when a key is presented — the caller's own reports. Every hit is a finished, cited, evidence-graded report. Returns {results:[{id,title,url}]}; pass an id to fetch. No key needed for the public corpus. (The research-connector contract: this is what a ChatGPT or Claude connector calls.)

輸入結構描述

{
  "type": "object",
  "properties": {
    "query": {
      "description": "Words to match against report titles and text.",
      "type": "string"
    }
  },
  "required": [
    "query"
  ]
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "results": {
      "items": {
        "additionalProperties": true,
        "properties": {
          "finished_at": {
            "type": [
              "string",
              "null"
            ]
          },
          "id": {
            "description": "Pass to fetch.",
            "type": "string"
          },
          "kind": {
            "description": "featured | free | own.",
            "type": "string"
          },
          "mode": {
            "type": "string"
          },
          "title": {
            "type": "string"
          },
          "url": {
            "description": "The public report page, or the API report URL for a private run.",
            "type": "string"
          }
        },
        "required": [
          "id",
          "title",
          "url"
        ],
        "type": "object"
      },
      "type": "array"
    }
  },
  "required": [
    "results"
  ],
  "additionalProperties": true
}
🟢fetch(id)

Read one report by the id search returned: the full markdown text with its citations and evidence header, plus title, url and metadata. Public reports need no key; a caller's own reports need theirs.

輸入結構描述

{
  "type": "object",
  "properties": {
    "id": {
      "description": "A result id from search (a run id).",
      "type": "string"
    }
  },
  "required": [
    "id"
  ]
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "id": {
      "type": "string"
    },
    "metadata": {
      "additionalProperties": true,
      "type": "object"
    },
    "text": {
      "description": "The report as markdown.",
      "type": "string"
    },
    "title": {
      "type": "string"
    },
    "url": {
      "type": "string"
    }
  },
  "required": [
    "id",
    "title",
    "text",
    "url"
  ],
  "additionalProperties": true
}
🟢route.classify(prompt)

Classify a freeform request into the right mode (research, news, finance, astrology, …) before spending a run. Never charged.

輸入結構描述

{
  "type": "object",
  "properties": {
    "prompt": {
      "description": "The freeform request to classify, e.g. 'compare the best espresso grinders under $300'.",
      "type": "string"
    }
  },
  "required": [
    "prompt"
  ]
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "alternates": {
      "description": "Other plausible modes, best first.",
      "type": "array"
    },
    "confidence": {
      "description": "Router confidence, 0–1.",
      "type": "number"
    },
    "fields": {
      "description": "Fields the router extracted from the prompt (mode-specific).",
      "type": "object"
    },
    "from_llm": {
      "description": "True when the LLM router answered, false for the offline heuristic.",
      "type": "boolean"
    },
    "mode": {
      "description": "The chosen mode id — pass it to runs.launch.",
      "type": "string"
    },
    "rationale": {
      "description": "Why this mode was picked.",
      "type": "string"
    }
  },
  "required": [
    "mode",
    "confidence"
  ],
  "additionalProperties": true
}
🟡runs.launch(input, max_workers, mode)

Launch a run: agents search the web, read sources, and write a cited report — real work, done for the caller. BILLED from the caller's credits. Returns the run record; poll runs.get until completed, then runs.report. Modes: research, news, product, trips, trainer, finance, astrology, horoscope (astrology/horoscope take Birth date constraint lines — see https://ioswarm.io/llms.txt).

輸入結構描述

{
  "type": "object",
  "properties": {
    "input": {
      "description": "The question or brief. astrology/horoscope take constraint lines ('- Birth date: YYYY-MM-DD', optional Birth time + Timezone, Latitude/Longitude) — contract at https://ioswarm.io/llms.txt.",
      "type": "string"
    },
    "max_workers": {
      "description": "Optional parallelism override (worker count). Omit for the mode default.",
      "type": "integer"
    },
    "mode": {
      "description": "A mode id from modes.list, e.g. 'research'.",
      "type": "string"
    }
  },
  "required": [
    "mode",
    "input"
  ]
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "estimated_credits": {
      "description": "The credit estimate reserved at launch; settled from real usage at completion.",
      "type": "number"
    },
    "run_id": {
      "description": "The run's id — pass to runs.get / runs.report / runs.cancel.",
      "type": "string"
    },
    "status": {
      "description": "'running' on acceptance.",
      "type": "string"
    }
  },
  "required": [
    "run_id",
    "status"
  ],
  "additionalProperties": true
}
🟢runs.get(run_id)

A run's status and metadata. Poll until status is completed/failed/cancelled.

輸入結構描述

{
  "type": "object",
  "properties": {
    "run_id": {
      "description": "The run_id returned by runs.launch.",
      "type": "string"
    }
  },
  "required": [
    "run_id"
  ]
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "cancelling": {
      "description": "A stop was requested but the run has not unwound yet.",
      "type": "boolean"
    },
    "error": {
      "description": "Failure detail when status is failed, else null.",
      "type": [
        "string",
        "object",
        "null"
      ]
    },
    "finished_at": {
      "description": "RFC 3339 completion time; null while running.",
      "type": [
        "string",
        "null"
      ]
    },
    "has_report": {
      "description": "True once runs.report will answer with the report.",
      "type": "boolean"
    },
    "meta": {
      "description": "Run metadata: id, mode, title, timing.",
      "type": "object"
    },
    "status": {
      "description": "running | completed | failed | cancelled.",
      "type": "string"
    },
    "usage": {
      "description": "Token meters for the run — null while the run is still running, an object once it settles.",
      "type": [
        "object",
        "null"
      ]
    }
  },
  "additionalProperties": true
}
🟢runs.wait(run_id, timeout_secs)

Block until a run finishes, then return its record — and its report when it completed. Runs take 3–10 minutes; call this instead of polling runs.get in a loop. Returns early with timed_out=true after timeout_secs (default 90, max 240): call it again with the same run_id to keep waiting. Never charged.

輸入結構描述

{
  "type": "object",
  "properties": {
    "run_id": {
      "description": "The run_id returned by runs.launch.",
      "type": "string"
    },
    "timeout_secs": {
      "description": "How long this call may wait before returning the current status. Default 90, max 240.",
      "type": "integer"
    }
  },
  "required": [
    "run_id"
  ]
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "report": {
      "description": "The report (as runs.report) when status is completed, else null.",
      "type": [
        "object",
        "null"
      ]
    },
    "run": {
      "description": "The run record (as runs.get).",
      "type": "object"
    },
    "run_id": {
      "type": "string"
    },
    "status": {
      "description": "running | completed | failed | cancelled.",
      "type": "string"
    },
    "timed_out": {
      "description": "True when the wait window elapsed first — call again.",
      "type": "boolean"
    },
    "waited_secs": {
      "type": "integer"
    }
  },
  "required": [
    "run_id",
    "status",
    "timed_out"
  ],
  "additionalProperties": true
}
🟢runs.report(run_id)

The finished, cited report for a completed run (report_text is markdown).

輸入結構描述

{
  "type": "object",
  "properties": {
    "run_id": {
      "description": "The run_id returned by runs.launch.",
      "type": "string"
    }
  },
  "required": [
    "run_id"
  ]
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "report_text": {
      "description": "The full report as markdown, citations inline.",
      "type": "string"
    }
  },
  "additionalProperties": true
}
🔴runs.cancel(run_id)

Stop a running job.

輸入結構描述

{
  "type": "object",
  "properties": {
    "run_id": {
      "description": "The run_id returned by runs.launch.",
      "type": "string"
    }
  },
  "required": [
    "run_id"
  ]
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "run_id": {
      "type": "string"
    },
    "status": {
      "description": "'cancelling' — the run unwinds itself; poll runs.get for the terminal state.",
      "type": "string"
    }
  },
  "required": [
    "run_id",
    "status"
  ],
  "additionalProperties": true
}
🟢billing.balance

The caller's credit balance and plan state.

輸入結構描述

{
  "type": "object",
  "properties": {}
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "balance": {
      "description": "Credits available right now.",
      "type": "number"
    },
    "currency": {
      "type": "string"
    },
    "granted_total": {
      "description": "Credits ever granted (welcome + top-ups).",
      "type": "number"
    },
    "plan": {
      "type": "string"
    },
    "spent_total": {
      "description": "Credits ever spent on runs.",
      "type": "number"
    },
    "user_id": {
      "type": "string"
    }
  },
  "required": [
    "balance"
  ],
  "additionalProperties": true
}
🟢billing.topup_terms

How to refill credits machine-to-machine: USDC on Base to a published address, verified on-chain. Returns the pay_to address, minimum, and rate. Public — no credentials needed.

輸入結構描述

{
  "type": "object",
  "properties": {}
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "asset": {
      "description": "'USDC'.",
      "type": "string"
    },
    "chain_id": {
      "description": "8453 (Base mainnet).",
      "type": "integer"
    },
    "contract": {
      "description": "The USDC token contract to transfer.",
      "type": "string"
    },
    "credits_per_usd": {
      "description": "Base rate; volume bonus applies on top.",
      "type": "number"
    },
    "enabled": {
      "description": "False while the operator has published no receiving address.",
      "type": "boolean"
    },
    "how": {
      "description": "The exact top-up procedure, step by step.",
      "type": "string"
    },
    "min_confirmations": {
      "description": "Confirmations required before crediting.",
      "type": "integer"
    },
    "min_usd": {
      "description": "Smallest top-up accepted, in USD.",
      "type": "number"
    },
    "network": {
      "description": "'base'.",
      "type": "string"
    },
    "pay_to": {
      "description": "The receiving address — present when enabled is true.",
      "type": "string"
    }
  },
  "required": [
    "enabled",
    "network",
    "asset",
    "contract",
    "min_usd",
    "credits_per_usd",
    "min_confirmations",
    "how"
  ],
  "additionalProperties": true
}
🟢modes.list

The launchable modes with a one-line contract for each. Static; never charged, no credentials needed.

輸入結構描述

{
  "type": "object",
  "properties": {}
}

輸出結構描述

{
  "type": "object",
  "properties": {
    "modes": {
      "items": {
        "additionalProperties": true,
        "properties": {
          "contract": {
            "description": "What the mode does, one line.",
            "type": "string"
          },
          "id": {
            "description": "Pass to runs.launch as mode.",
            "type": "string"
          }
        },
        "required": [
          "id",
          "contract"
        ],
        "type": "object"
      },
      "type": "array"
    },
    "route_first": {
      "description": "How to classify a freeform prompt into a mode id.",
      "type": "string"
    }
  },
  "required": [
    "modes"
  ],
  "additionalProperties": true
}

建議的提示詞

search_research
Search for information about [topic] using SwarmIO
預期的工具： search
find_specific
Find [specific item] using SwarmIO
預期的工具： search
retrieve_data
Get details about [item] from SwarmIO
預期的工具： fetch
fetch_info
Fetch [information type] using SwarmIO
預期的工具： fetch
list_items
List all [items] available in SwarmIO
預期的工具： modes.list

社群

為此伺服器評分

證據

近期觀測

已驗證未記錄版本12 個工具
已驗證未記錄版本12 個工具