Countly

# Fix list: Countly

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/countly, the October 2026 research run, assessed 8 October 2026. Grade C, 54.8 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 Countly: 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, 28 out of 100, up to 14.4 more on the total

Why it scored 28: Graded on the hosted lines, for a Countly Flex server reached through its HTTP API and the official MCP server. No status page was found on countly.com, the Flex page or the docs, and status.countly.com did not resolve (0). With no page there is no readable incident history (5). The MCP server's HTTP endpoint is limited to 120 requests a minute per IP, but the Countly server's own `api_rate_limit_requests` setting ships at 0 (off), and the Flex terms reserve usage limits without numbers (8 of 15). No 429 or retry guidance and no idempotency keys were found in the API reference (0). The pricing page lists an SLA as optional on Flex and included in Enterprise, with no SLA text published (5 of 10). The API and MCP server 1.7.0 are generally available (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. Payments & pricing, 17 out of 100, up to 10.4 more on the total

Why it scored 17: No x402, MPP or L402 (0). The pricing page gives only a starting price for Flex, $175 a month, priced on monthly active users and data points, with tier prices shown during signup and Enterprise on quotation (7 of 20). The 14-day Flex trial needs a credit card. Countly Lite is free to self-host with no card, without funnels, retention, cohorts or Drill, so half credit (10). A person signs up in a browser, creates a server and makes a token in the dashboard (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.

## 3. Security & auth, 63 out of 100, up to 6.5 more on the total

Why it scored 63: Auth tokens are scoped to apps and endpoint patterns, expire after a set lifetime, are revocable and carry create, read, update and delete permissions since 25.03.54. The dashboard `api_key` has its user's full access (26 of 30). Less 10 because `api_key` and `auth_token` are documented as query-string parameters, with the `countly-token` header as the recommended alternative (16). MCP read-only mode and per-category CRUD levels, bounded by the token's permissions. No confirmation step for destructive tools (16). Tools return event names, user properties and crash data written by end users, and no prompt-injection guidance was found. The docs suggest a self-hosted model for sensitive data (3). A system logs API records dashboard actions, and the security FAQ cites more than 30 log types (12). No security.txt. SECURITY.md gives a disclosure address and a bounty for medium severity and above, security fixes are listed in the public changelog, and the site claims ISO 27001, 27017, 27018 and SOC 2 Type II without public reports (16).

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.

## 4. Agent ergonomics, 66 out of 100, up to 5.5 more on the total

Why it scored 66: The MCP README counts 209 tools in 43 categories, far past 30 (5). Add back 8 because the server hides tools the edition, enabled plugins or user's permissions cannot run, and `COUNTLY_TOOLS_*` settings switch categories off (13). List endpoints take `iDisplayStart` and `iDisplayLength`, and MCP list tools take `skip`, `limit` and `search` (15). API errors are `{"result": "message"}` strings with HTTP status codes and no error codes. The MCP server returns failures as `isError` results that name the missing plugin or permission (12). MCP tools carry readOnlyHint, destructiveHint and idempotentHint since 1.7.0. No idempotency keys on API writes (14). MCP tools accept `app_name` in place of `app_id` and set defaults. The SDKs cover data collection, not the read API (12).

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. Schema & documentation, 75 out of 100, up to 4.1 more on the total

Why it scored 75: No OpenAPI document was found on api.count.ly or in Countly/countly-server. Every MCP tool has a typed JSON Schema input (18 of 25). countly.com/llms.txt indexes 366 help articles and points automated clients to the help centre's JSON API, while api.count.ly has no llms.txt and article pages answered 403 to our reader (8). MCP tool descriptions name the endpoint, the plugin required and the tool to use instead, and API pages state permissions and behaviour (16). 82 enums across the tool sources, but the HTTP API takes JSON arrays and objects as query-string values (9). API pages carry request examples, response fields and error responses, as plain message strings (11). The API reference is versioned 24.05, 25.03 and 26.01, and the server and MCP server keep public changelogs (13).

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.

## 6. Transparency & trust, 70 out of 100, up to 2.6 more on the total

Made of editorial 64, provenance 76.

Why it scored 70: Countly Lite is AGPL-3.0 with added branding terms under Section 7, the MCP server is MIT, and the Enterprise plugins that hold funnels, retention, cohorts and Drill are closed (20 of 30). The terms of 25 September 2026 cover Flex and incorporate a public DPA. The privacy policy says it does not apply to data customers process and gives no retention periods. Granular data retention is chosen when a Flex server is created (20). No deprecation policy was found. The MCP README says URL token parameters will be removed in a future release, with no date (6 of 20). Sub-processors are listed per product with locations, with 30 days' notice of changes, and Flex customers choose a hosting region. MCP usage reporting to stats.count.ly is disclosed, on by default, with an opt-out (18).

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):

- 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)
- security.txt: not found (0 of 10)

## 7. Maintenance & community, 80 out of 100, up to 1.8 more on the total

Why it scored 80: `countly-mcp-server` 1.7.0 was published on 7 October 2026 and server tag 25.03.54 on 22 September 2026 (30). Four server tags (25.03.49, .50, .51, .54) and MCP 1.5.0 and 1.7.0 since 10 July 2026 (20). 821 commits on the server's default branch since 10 July 2026, a Discord community and a support portal. We could not read issue response times because the GitHub API refused us (12 of 25). SDK docs are at 26.1.x for Android, iOS, Web and Flutter, while the NodeJS SDK docs are at 24.10.x. The MCP server is not in the official MCP registry (10 of 15). CI, CodeQL and Docker workflows on the server, and test, nightly end-to-end and CodeQL workflows on the MCP server (8).

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.

## 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: GitHub stars, open issues, response times and published security advisories for Countly/countly-server. The GitHub API answered with a rate limit, so we read the repositories by clone only
- unchecked: help centre article pages on support.countly.com, which answered 403 with a bot check. We read five articles through the help centre's public JSON API, which countly.com/llms.txt names for automated clients, and left the rest unread
- unchecked: Flex tier prices, limits on monthly active users and data points, and regional price differences, which appear only inside the signup flow
- unchecked: the domain a Flex server answers on by default. The MCP README's example is https://your-server.count.ly
- unchecked: the text of the Flex and Enterprise SLA, and the certification reports behind the ISO 27001, 27017, 27018 and SOC 2 Type II claims
- unchecked: the live `tools/list` of the MCP server, which needs a Countly server and token. The count of 209 tools is the README's
- The help centre says Funnels, Retention, Flows, Cohorts and Drill are built into Flex, while the pricing page places funnels, retention and flows in the Intelligence capability, a Flex add-on. We followed the pricing page and made no deduction
- The server changelog for 25.03.54 (22 September 2026) lists 20 security fixes, among them authorisation checks on alerts, reports and widget endpoints. No incident was found, and no deduction was made
- countly.com/llms.txt opens with a note addressed to LLMs and AI agents saying the file is authoritative where third-party sources disagree. We treated it as vendor material and made no deduction
- mcp.count.ly reported version 1.5.0 on 8 October 2026, one release behind 1.7.0, whose changelog lists five security fixes
- The privacy policy excludes data customers process with the Services, so the DPA is the document that covers event data
- The lead's interface (HTTP APIs and SDKs) was right but left out the official MCP server

## Weaknesses

- No public status page or incident history was found for Flex servers or for mcp.count.ly
- No OpenAPI document was found, and the server's API rate limit is a setting that ships switched off with no published value for Flex
- The API accepts `api_key` and `auth_token` in the URL query string, and the `api_key` has its user's full read and write access
- Flex starts at $175 a month, its 14-day trial needs a credit card, and tier prices appear only during signup
- Funnels, retention, flows, cohorts and Drill need Enterprise or a paid Flex capability, and are absent from Countly Lite
- The MCP server reports usage to stats.count.ly by default until `ENABLE_ANALYTICS=false` is set

## 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.

- Create a dedicated token in the Token Manager, limited to the apps and endpoints the task needs, and send it in the `countly-token` header, not the URL
- Set `COUNTLY_TOOLS_ALL=R` unless the task needs writes. The token's own permissions still apply and the stricter of the two wins
- Call `apps_list` first. Most tools need `app_id` or an exact `app_name`
- Check `get_version` for the detected edition before planning funnel, retention or cohort queries, which need Enterprise plugins
- Set `ENABLE_ANALYTICS=false` to stop usage reports to stats.count.ly, and prefer the stdio server to mcp.count.ly so the server URL and token stay local

## 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.