# SwarmIO (remote · ioswarm.io)

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

- Trust score: 64/100 (medium)
- Change this week: 0
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-09-22

## Components

- remote · `ioswarm.io`: 64/100 (this document), [markdown](https://verifymcp.io/servers/io-ioswarm-swarmio/ioswarm.md), [page](https://verifymcp.io/servers/io-ioswarm-swarmio/ioswarm)

## Channel facts

- Endpoint: `https://ioswarm.io/mcp`
- Transports: `streamable-http`
- Auth: `required`
- Version: `1.0.1`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically against the live server, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-09-22.

- **Endpoint Security**: 57/100
  - The endpoint's TLS certificate is valid, in date, and uses a strong key.
  - Authorisation not fully verified: this server exposes a tool marked destructive (runs.cancel) and its handshake is open, but we could not confirm whether a tool call is gated, so we do not assert it is callable unauthenticated.
  - HTTPS is enforced; there's no plaintext access path.
  - HSTS check failed: the Strict-Transport-Security header is absent.
  - DNSSEC check failed: this domain isn't protected by DNSSEC.
- **Transport & Reachability**: 100/100
  - Verified streamable-http transport via a live MCP handshake.
- **Schema Quality & AI Usability**: 82/100
  - AI-judged instruction clarity (excellent).
  - Tool/resource definitions use about 1057 tokens (~88/item across 12 items; 12 tools + 0 resources), lean.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 0/100
  - Stability check failed: schema churn in the 26 days we've observed: 8 tool removals, 0 breaking changes, 0 auth/transport breaks, 12 additions.
- **Tool Coverage**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
  - Structured output schemas are declared (100% of tools); any adoption earns full credit.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - We read all 12 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 13 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 60/100
  - Spec-recency check failed: implements MCP spec 2025-06-18; the latest is 2026-07-28.

## Install

### How do I install the SwarmIO MCP server?

SwarmIO is a hosted endpoint at https://ioswarm.io/mcp, so there is nothing to install locally. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add --transport http io-ioswarm-swarmio 'https://ioswarm.io/mcp'
```

### Cursor

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

### VS Code

```json
{
  "servers": {
    "io-ioswarm-swarmio": {
      "type": "http",
      "url": "https://ioswarm.io/mcp"
    }
  }
}
```

### Codex

```toml
[mcp_servers.io-ioswarm-swarmio]
url = "https://ioswarm.io/mcp"
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "io-ioswarm-swarmio": {
      "type": "remote",
      "url": "https://ioswarm.io/mcp",
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add io-ioswarm-swarmio --url 'https://ioswarm.io/mcp' --transport streamable-http
```

### Hermes

```yaml
mcp_servers:
  io-ioswarm-swarmio:
    url: "https://ioswarm.io/mcp"
```

### Netclaw

```json
{
  "McpServers": {
    "io-ioswarm-swarmio": {
      "Transport": "http",
      "Url": "https://ioswarm.io/mcp"
    }
  }
}
```

### Vellum

```bash
assistant mcp add io-ioswarm-swarmio -t streamable-http -u 'https://ioswarm.io/mcp'
```

### Other

```json
{
  "mcpServers": {
    "io-ioswarm-swarmio": {
      "type": "http",
      "url": "https://ioswarm.io/mcp"
    }
  }
}
```

The mcpServers block is a cross-client convention. Remote transports vary, so check your client's docs.

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-09-17 (score 64, 0)

- [security] The server rewrote its instructions, which are the text every model session reads
- [functional regression] Schema quality: 69 → 88
- [functional] New tool “fetch”
- [functional] New tool “market.pulse”
- [functional] New tool “runs.wait”
- [functional] New tool “search”

### 2026-08-28 (score 64, +4)

- [security regression] Stability: unverified → fail
- [security regression] A breaking change shipped without a version bump: still 0.1.0
- [security regression] Tool “cancel_run” was removed
- [security regression] Tool “get_balance” was removed
- [security regression] Tool “get_report” was removed
- [security regression] Tool “get_run” was removed
- [security regression] Tool “get_topup_terms” was removed
- [security regression] Tool “launch_run” was removed
- [security regression] Tool “list_modes” was removed
- [security regression] Tool “route_prompt” was removed
- [security] The server rewrote its instructions, which are the text every model session reads
- [security] New tool “runs.cancel”, which the server declares destructive
- [functional regression] Schema quality: 52 → 69
- [functional improvement] Tool coverage: 29% → 100%
- [functional] First check of Tool coverage: 100
- [functional] Schema quality: good → excellent
- [functional] New tool “billing.balance”
- [functional] New tool “billing.topup_terms”
- [functional] New tool “modes.list”
- [functional] New tool “route.classify”
- [functional] New tool “runs.get”
- [functional] New tool “runs.launch”
- [functional] New tool “runs.report”

### 2026-08-27 (score 60)

First indexed and scored.

## MCP tools (12)

### `market.pulse` (~94 tokens)

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.

Output parameters:

- `as_of` (string): RFC 3339 time this document was rendered.
- `attribution` (string)
- `free` (boolean): Always true — this call is never charged.
- `news` (object)
- `next` (object): How to get the same data over MCP and how to commission a real report.
- `prices` (array): One entry per symbol that priced this cycle, in list order.
- `prices_as_of` (string|null)
- `symbols` (array): The fixed symbol list served.

### `search` (~114 tokens)

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.)

Input parameters:

- `query` (string, required): Words to match against report titles and text.

Output parameters:

- `results` (array)

### `fetch` (~64 tokens)

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.

Input parameters:

- `id` (string, required): A result id from search (a run id).

Output parameters:

- `id` (string)
- `metadata` (object)
- `text` (string): The report as markdown.
- `title` (string)
- `url` (string)

### `route.classify` (~64 tokens)

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

Input parameters:

- `prompt` (string, required): The freeform request to classify, e.g. 'compare the best espresso grinders under $300'.

Output parameters:

- `alternates` (array): Other plausible modes, best first.
- `confidence` (number): Router confidence, 0–1.
- `fields` (object): Fields the router extracted from the prompt (mode-specific).
- `from_llm` (boolean): True when the LLM router answered, false for the offline heuristic.
- `mode` (string): The chosen mode id — pass it to runs.launch.
- `rationale` (string): Why this mode was picked.

### `runs.launch` (~190 tokens)

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).

Input parameters:

- `input` (string, required): 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.
- `max_workers` (integer): Optional parallelism override (worker count). Omit for the mode default.
- `mode` (string, required): A mode id from modes.list, e.g. 'research'.

Output parameters:

- `estimated_credits` (number): The credit estimate reserved at launch; settled from real usage at completion.
- `run_id` (string): The run's id — pass to runs.get / runs.report / runs.cancel.
- `status` (string): 'running' on acceptance.

### `runs.get` (~41 tokens)

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

Input parameters:

- `run_id` (string, required): The run_id returned by runs.launch.

Output parameters:

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

### `runs.wait` (~120 tokens)

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.

Input parameters:

- `run_id` (string, required): The run_id returned by runs.launch.
- `timeout_secs` (integer): How long this call may wait before returning the current status. Default 90, max 240.

Output parameters:

- `report` (object|null): The report (as runs.report) when status is completed, else null.
- `run` (object): The run record (as runs.get).
- `run_id` (string)
- `status` (string): running | completed | failed | cancelled.
- `timed_out` (boolean): True when the wait window elapsed first — call again.
- `waited_secs` (integer)

### `runs.report` (~38 tokens)

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

Input parameters:

- `run_id` (string, required): The run_id returned by runs.launch.

Output parameters:

- `report_text` (string): The full report as markdown, citations inline.

### `runs.cancel` (~28 tokens)

Stop a running job.

Input parameters:

- `run_id` (string, required): The run_id returned by runs.launch.

Output parameters:

- `run_id` (string)
- `status` (string): 'cancelling' — the run unwinds itself; poll runs.get for the terminal state.

### `billing.balance` (~17 tokens)

The caller's credit balance and plan state.

Output parameters:

- `balance` (number): Credits available right now.
- `currency` (string)
- `granted_total` (number): Credits ever granted (welcome + top-ups).
- `plan` (string)
- `spent_total` (number): Credits ever spent on runs.
- `user_id` (string)

### `billing.topup_terms` (~48 tokens)

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.

Output parameters:

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

### `modes.list` (~29 tokens)

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

Output parameters:

- `modes` (array)
- `route_first` (string): How to classify a freeform prompt into a mode id.

## Diagnostics

Captured diagnostic sections: TLS, DNSSEC, Authorisation, Transports. The full working is on the page: https://verifymcp.io/servers/io-ioswarm-swarmio/ioswarm#diagnostics

## Score history

- 2026-09-22: 64
- 2026-09-21: 64
- 2026-09-20: 64
- 2026-09-19: 64
- 2026-09-18: 64
- 2026-09-17: 64
- 2026-09-16: 64
- 2026-09-15: 64
- 2026-09-14: 64
- 2026-09-13: 64
- 2026-09-12: 64
- 2026-09-11: 64
- 2026-09-10: 64
- 2026-09-09: 64
- 2026-09-08: 64
- 2026-09-07: 64
- 2026-09-06: 64
- 2026-09-05: 64
- 2026-09-04: 64
- 2026-09-03: 64
- 2026-09-02: 64
- 2026-09-01: 64
- 2026-08-31: 64
- 2026-08-30: 64
- 2026-08-29: 64
- 2026-08-28: 64
- 2026-08-27: 60

## Common questions

### What is the SwarmIO MCP server?

SwarmIO is an MCP server listed in the public MCP registry as io.ioswarm/swarmio. AI that does real work for you: cited web research, plans, finance, horoscopes. Self-serve. This page covers its hosted endpoint (https://ioswarm.io/mcp).

### Is the SwarmIO MCP server safe to use?

SwarmIO scores 64 out of 100 on VerifyMCP. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the SwarmIO MCP server expose?

SwarmIO exposes 12 tools: market.pulse, search, fetch, route.classify, runs.launch, and 7 more. Their descriptions and schemas cost roughly 847 tokens of context every time the server is loaded.

### Does the SwarmIO MCP server require authentication?

No. We connected to SwarmIO without credentials and it answered, so anything it exposes is reachable by anyone who knows the address.

### Is the SwarmIO MCP server still maintained?

SwarmIO is still listed as active in the MCP registry. We last reached this channel on 22 September 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

## Links

- Remote endpoint: https://ioswarm.io/mcp
- Authorisation metadata: https://ioswarm.io/.well-known/oauth-protected-resource/mcp
- Website: https://ioswarm.io/
- Changelog RSS feed: https://verifymcp.io/servers/io-ioswarm-swarmio/ioswarm.xml
- Changelog JSON feed: https://verifymcp.io/servers/io-ioswarm-swarmio/ioswarm.json
- HTML version of this page: https://verifymcp.io/servers/io-ioswarm-swarmio/ioswarm