Fuse MCP

# Fix list: Fuse MCP

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/fuse-mcp, the October 2026 research run, assessed 8 October 2026. Grade D, 47.9 out of 100.

This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public.

For a coding agent working on Fuse MCP: work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published.

## 1. Reliability, 40 out of 100, up to 12 more on the total

Why it scored 40: Graded as a hosted service. No status page is linked from the site, the docs or the sitemap, and status.fuseai.com did not connect from our network (0). No readable incident history (5). MCP limits are published at 60 requests a minute and 5,000 a day per user (15). The server answers 429 with `RATE_LIMIT_EXCEEDED` and the docs say to back off and retry, and long jobs return at once and are polled. No `Retry-After` is documented for the MCP server, there are no idempotency keys, and the docs say changes have no undo (10 of 15). No SLA found (0). The server carries no beta or preview label, although its tool list was replaced in October 2026 (10).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):

Hosted APIs, MCP servers, models and platforms.

- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).
- 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so.
- 15, rate limits documented with numbers.
- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.
- 10, an SLA published for any paid tier.
- 10, the surface agents use is generally available, not beta or preview.

Local packages, SDKs, frameworks and stdio MCP servers.

- 20, installs from an official package with supported runtimes stated.
- 25, a public CI and test suite, passing on the default branch.
- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).
- 15, semver discipline and breaking changes called out in a changelog.
- 15, version 1.0 or later, or declared stable.

Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.

## 2. Security & auth, 45 out of 100, up to 9.6 more on the total

Why it scored 45: OAuth authorisation code with PKCE and dynamic client registration, one-hour access tokens and rotating refresh tokens. The only scope is `mcp:tools`, so every connection has the user's full permissions (22 of 30). There is no read-only mode. Destructive tool descriptions tell the model to get the user's confirmation and launch tools tell it to confirm first, but the server enforces none of this and changes apply at once. Billing, API key and team-removal tools are left out, as are tools that take a password or key (8 of 20). Tools return inbox email, LinkedIn messages and scraped web pages. The only guidance found is to connect trusted clients and keep approval prompts on (4 of 15). Audit logging is listed for Enterprise on the pricing page, and a credit usage ledger exists. No per-call log for MCP is documented (4 of 15). The home page lists SOC 2 Type I and CASA Tier 2 with no report or trust centre found. No security.txt, disclosure policy or bug bounty found (7 of 20).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-security):

- 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option.
- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.
- 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10.
- 0 to 15, audit logs or per-call visibility for the operator.
- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.

Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing.

## 3. Payments & pricing, 30 out of 100, up to 8.8 more on the total

Why it scored 30: No x402, MPP or L402 (0). Plan prices and the credit cost of each action are public, and `get_credit_prices` returns them. Top-up prices are shown only in the app and the Free allowance is stated two ways (15 of 20). A Free plan exists at $0. The Solo plan card is where MCP is listed, and we could not confirm that Free includes it or that signup needs no card (15 of 20). A person signs in through a browser to approve each connection (0).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):

The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).

- 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which.
- 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login.
- 20, a free tier or trial that doesn't need a card.
- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).

Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied.

Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol.

## 4. Agent ergonomics, 57 out of 100, up to 7 more on the total

Why it scored 57: 290 tools and about half a megabyte of definitions, with no toolsets, dynamic loading or read-only subset. The docs warn that some clients pass only the first 40 tools (5 of 25). List reads return a `data` array with `pagination`, searches return a scroll token or cursor, and tools take limits and filters (18 of 20). Rejected calls set `isError` and return codes such as `TOOL_ARGUMENT_VALIDATION_FAILED` with the missing field named. Other failures are short plain text (15 of 20). The docs say every tool carries `readOnlyHint` or `destructiveHint`, which we could not read from the server. No idempotency keys, and sends are not marked destructive (14 of 20). No official SDK found, which matters less for an MCP server, and most changes need an id looked up first (5 of 15).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):

- 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries).
- 20, pagination, filtering and output-size controls.
- 20, actionable, documented error responses, codes and messages an agent can recover from.
- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.
- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.

Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs.

## 5. Maintenance & community, 35 out of 100, up to 5.7 more on the total

Why it scored 35: The tool reference is dated 7 October 2026 and the tool list was rebuilt in October 2026 (30). No changelog or release notes found, so three dated entries in 90 days could not be shown (0). No public changelog or issue tracker. Support is by email at support@fuseai.com, with Docs and community on Free and email support on paid plans. We did not test it (5 of 15). Not in the official MCP registry. Searches for `fuse` and `tryfuse` returned none of the vendor's, and the `fuseai` search timed out twice. No official SDK found (0). No package to assess (0 of 10).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):

- 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older.
- 20, at least three releases or dated changelog entries in the last 90 days.
- 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15.
- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).
- 10, package health, current dependencies and CI.

Models are read for deprecation notice periods and model churn rather than release counts.

## 6. Schema & documentation, 72 out of 100, up to 4.6 more on the total

Why it scored 72: The docs say every tool ships a description and an input schema, about half a megabyte in all. We could not call tools/list without an account, so the schemas themselves were not read. The REST API beside it has a public OpenAPI 3.1 spec with 146 operations (22 of 25). llms.txt, Markdown twins of every page and a skill file (10). The tool reference shows descriptions that state purpose, the preferred tool for a job and when a destructive tool may be called (16 of 20). Enum lookup tools such as `get_company_search_enums` exist, but input constraints were not read first-hand (9 of 15). Error results are documented with codes and two JSON examples, plus four plain-text failures. The docs say the codes do not always match the REST catalogue (12 of 15). No changelog was found and the docs say the server does not announce changes to the tool list. The tool reference is dated 7 October 2026 (3 of 15).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):

APIs and MCP servers.

- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).
- 10, llms.txt or Markdown docs served for agents.
- 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference.
- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.
- 0 to 15, examples and documented error responses.
- 15, versioning and a public changelog.

Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference.

## 7. Transparency & trust, 49 out of 100, up to 4.5 more on the total

Made of editorial 36, provenance 62.

Why it scored 49: Closed service with published terms of service, last updated 1 July 2026, naming FuseAI, LLC (15). The privacy policy of the same date is general. Retention is as long as necessary with no periods, and no DPA or subprocessor list was found. It says Google Workspace data is not used to train AI models and points to a Do Not Sell or Share tool (10 of 30). No deprecation policy. The docs record the October 2026 removal of the earlier tools and that the earlier server URL keeps working (5 of 20). The docs name 39 external data providers. Hosting, data locations and subprocessors are not disclosed (6 of 20).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):

- 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms.
- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).
- 0 to 20, a deprecation policy or notices with dates.
- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).

The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.

Provenance checks not met in full (half of this category, computed from checked facts):

- Domain age: fuseai.com, registered 2017-03-11 (9 years) (11 of 15)
- Terms of service: read, states 5 of the 7 things a reader expects, and has 1 clause that costs points (6.3 of 10)
- Privacy policy: read, states 7 of the 8 things a reader expects (9.3 of 10)
- Status page: not found (0 of 10)
- Changelog: not found (0 of 10)
- security.txt: not found (0 of 10)

## What we couldn't check

What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it.

- unchecked: the server's own tools/list, so the tool count, input schemas and annotations are as the vendor's docs state them
- unchecked: whether the Free plan includes MCP access and whether signup asks for a card. Signup was not attempted
- unchecked: status.fuseai.com, status.tryfuse.ai and trust.fuseai.com did not connect from our network, and no page links to them
- unchecked: the official MCP registry search for `fuseai` timed out twice. Searches for `fuse` and `tryfuse` returned nothing of the vendor's
- Whether customers were told before the earlier MCP tools were removed in October 2026. No announcement or changelog was found, and no deduction was taken
- The vendor's 1,000+ figure on the home page counts customer companies, not integrations. No page supports a claim of 1,000+ integrations
- The Free plan allowance (24,000 credits a year on the plan card, 2,000 in the usage table and on the machine-readable page) and the annual discount (20 or 25 per cent) differ between vendor pages
- Which billing period the $60 and $200 monthly prices apply to, since the page also shows a 20 per cent annual discount

## Weaknesses

- The only OAuth scope is `mcp:tools`, so a connection can do everything its user can, including sending email and LinkedIn messages
- The 290 tools come to about half a megabyte of descriptions and schemas, with no toolsets or read-only subset
- The October 2026 rebuild removed the earlier tools such as `prospect_search`, and the docs say the server does not announce tool list changes
- No status page, changelog, SLA, security.txt or disclosure contact was found on the site or in the docs
- Vendor pages disagree on the Free allowance (24,000 credits a year or 2,000) and on the provider count (40+, 39 or 20+)

## What costs an agent a turn today

The notes we give agents before they call it. Each one is a workaround an agent shouldn't need.

- Keep the client's approval prompt on for every tool that is not read-only. Sends, CRM pushes and Slack posts run at once with no undo
- Launch a campaign only on an explicit request. Creating one makes a draft, and `update_campaign_status_tool` starts real outreach
- Call `get_credits_balance` and `get_credit_prices` first. Every call needs at least 100 credits, and searches cost 2 credits a result
- Resolve filter values with `lookup_filter_values` or `get_company_search_enums` before `search_people` or `search_companies`
- Start long work with `deep_research_start`, then poll `deep_research_status`. A tool call has 60 seconds, and runs take 5 to 20 minutes

## When it's done

Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.