Rootly MCP Server

# Fix list: Rootly MCP Server

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/rootly-mcp, the October 2026 research run, assessed 9 October 2026. Grade C, 59.7 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 Rootly MCP Server: 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, 48 out of 100, up to 10.4 more on the total

Why it scored 48: Scored as a hosted MCP server, the path the docs recommend. status.rootly.com is linked from the site and llms.txt, but it answered our reader with a bot check, so its components (15) and incident history (5) were unread. The REST API behind the tools allows 3,000 read calls and 3,000 write calls a minute per API key, and 50 alert creations a minute. No separate limit for mcp.rootly.com was found (12). A 429 returns a JSON error with `X-RateLimit-Limit`, `X-RateLimit-Remaining`, `X-RateLimit-Used` and `X-RateLimit-Reset` headers. No `Retry-After`, backoff guidance or idempotency keys were found (8). The pricing table lists "99.99 % Reliability" for on-call, and no SLA document was found (0). The hosted `/mcp` and `/sse` endpoints carry no beta label. The README calls Code Mode experimental while the docs recommend it (8).

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, 20 out of 100, up to 10 more on the total

Why it scored 20: No x402, MPP or L402 (0). Plan prices are public. Essentials is $20 per product per user a month, with MCP and API access included, and Enterprise is by quote. No per-call price (10). No free tier. Self-serve sign-up starts a trial of about two weeks per the pricing page. The sign-up page answered with a bot check, so whether a card is needed wasn't established, and the line takes half marks (10). A person signs up in a browser and either logs in through OAuth or creates an API key (0). The package is free to run but needs a Rootly account, so the self-hosted rule doesn't apply.

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, 65 out of 100, up to 6.1 more on the total

Why it scored 65: OAuth 2 with dynamic client registration, PKCE with S256, one-hour access tokens, rotating refresh tokens, a revocation endpoint and `:read` and `:write` scopes per resource. The MCP server's resource metadata advertises a single `all` scope since 2.3.13, so MCP logins aren't narrowed by scope. API keys are global with a chosen role, team or personal, sent in the `Authorization` header (26). Delete operations are off by default and API key, user, role and webhook endpoints are excluded. Write tools are on by default for hosted connections, the documented read-only switch applies to servers the owner runs, and no confirmation step was found (11). Incident text, alert payloads and meeting transcripts are untrusted content, and no injection guidance was found (3). `list_audits` reads Rootly's audit log, which the pricing page lists under Enterprise (9). security.txt valid to 1 September 2027, a disclosure policy with safe harbour and response times, no paid bounty, SOC 2 Type II stated with the report on request, and dependency advisories recorded in the changelog (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, 77 out of 100, up to 3.7 more on the total

Why it scored 77: The default endpoint lists about 218 tools per the README, which the checklist scores at 5. A slim profile (77 names in the source, about 70 per the README), a Code Mode endpoint with five meta-tools and an exact allowlist for servers the owner runs add back 10 (15). List tools page with bounded sizes, strip alert and incident lists to summary fields and clamp out-of-range counts, reporting each change under `argument_adjustments` (18). Tool errors return `error_type` and a message, with hints for plan-gated 404s, the pagination cap and replaced alert routing tools (16). Every tool sets `readOnlyHint`, `destructiveHint` and `openWorldHint` since 2.3.19, and writes set `idempotentHint`. `create_incident` has no idempotency key (15). Curated tools have no required parameters beyond record references and accept common aliases such as `limit`. Rootly lists REST SDKs for Go, Python and other languages (13).

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, 78 out of 100, up to 3.6 more on the total

Why it scored 78: Read from the source at v2.3.21. Curated tools declare typed pydantic inputs and the rest are generated from Rootly's OpenAPI 3.0.1 description, 565 operations in the bundled copy, which the API docs also link for download (25). rootly.com/llms.txt and Markdown twins of the docs pages (10). Curated tools such as `list_incidents` say when to use them and when to prefer `search_incidents`. Generated tools carry the API's short summaries, for example "List alerts" (11). Sort orders use enums and generated tools keep the API's `include` enums, but severity, status and ID lists are free strings, with IDs passed comma-separated (9). The README and docs give worked calls for five tools and a troubleshooting section. The error types the server returns aren't documented (8). Semantic versioning with a dated Keep a Changelog file. The docs page still names two tools in camelCase and links the old Rootly-AI-Labs repository (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.

## 6. Transparency & trust, 80 out of 100, up to 1.8 more on the total

Made of editorial 65, provenance 95.

Why it scored 80: The server is Apache-2.0 and the hosted service runs under Rootly's Terms of Use of July 2026 (27). A privacy policy of October 2026, a subprocessor list and AI data pages are public, and retention is stated only in general terms. The README says hosted telemetry records tool arguments, responses, timing and errors through AgentCat with redaction hooks. AgentCat isn't on the subprocessor list and the MCP docs page doesn't mention telemetry (15). No deprecation policy was found. camelCase tool names stay callable as hidden aliases with no removal date, and the terms say notice of changes isn't always practical (8). The subprocessor list names about 30 processors with purpose, data and location, nearly all in the United States (15).

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 6 of the 7 things a reader expects, and has 2 clauses that cost points (5.1 of 10)

## 7. Maintenance & community, 81 out of 100, up to 1.7 more on the total

Why it scored 81: v2.3.21 was tagged on 5 October 2026, four days before this check (30). Six dated releases between 23 July and 5 October 2026 (20). The commit log shows pull requests merged through 6 October and a fix credited to an outside contributor. We couldn't read the issue tracker, because api.github.com didn't answer (12). The official MCP registry lists `com.rootly/mcp-server` under Rootly's own domain namespace, but the entry is 2.2.4 from 18 February 2026 and names only the SSE endpoint (10). CI runs lint, two type checkers and tests on Python 3.12 and 3.13, with a lockfile check, Dependabot and a dependency review workflow (9).

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.

## Deductions

Each comes off the total. A fixed and documented problem counts for less at the next check.

- Between 23 and 29 September 2026 two changes to hosted telemetry broke tool calls. Removing an injected `context` parameter failed fifteen curated tools for clients with cached tool lists, and a PostHog analytics layer failed every hosted tool call until it was switched off. Both are fixed and described in the 2.3.21 changelog, so 3 points (https://github.com/rootlyhq/rootly-mcp-server/blob/main/CHANGELOG.md)

## 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: status.rootly.com components and incident history. The host answered its robots.txt request with a Cloudflare bot check (HTTP 403), so we treated it as closed and didn't retry
- unchecked: whether the free trial needs a card. rootly.com/users/sign_up answered with a bot check
- unchecked: the tool list the hosted endpoint serves. It needs a login, so tool counts and annotations come from the source and README at v2.3.21
- unchecked: GitHub stars and the issue tracker. api.github.com didn't answer our robots.txt request, so we treated it as closed
- unchecked: PyPI downloads, because PyPI's robots.txt closes the paths that hold them
- Where AgentCat stores hosted telemetry, and whether Rootly runs it or a third party does, wasn't established. It isn't on the subprocessor list
- How long hosted tool calls failed in late September 2026 isn't stated in the changelog. The dates are those of the merged fixes
- Whether hosted connections can be made read-only other than by the role on a Global API key wasn't established. The OAuth docs list `:read` scopes, but the MCP resource metadata advertises only `all`
- The Terms of Use bar crawling or scraping the Services by manual or automated means. Recorded as a fact with no deduction, and it matters before any probe is run
- rootly.com/llms.txt and the docs pages carry text addressed to AI models (when to recommend Rootly, and which index to fetch first). We treated it as data and didn't act on it
- The lead was right on the endpoints, licence and repository move. It didn't say the docs recommend Code Mode, or that the registry entry is stale at 2.2.4
- mcp.rootly.com and registry.modelcontextprotocol.io answered 404 for robots.txt, so their documented pages were read. security.rootly.com allows the page we read

## Weaknesses

- Hosted connections expose write tools by default, and the docs describe the read-only switch only for servers the owner runs
- The 2.3.21 changelog records two hosted regressions in late September 2026, one failing fifteen tools for clients with cached schemas and one failing every tool call
- Hosted telemetry records tool arguments and responses through AgentCat per the README. AgentCat isn't named on Rootly's subprocessor list
- status.rootly.com answered our reader with a bot check, so its components and incident history were unread
- The official registry entry `com.rootly/mcp-server` is still 2.2.4 from 18 February 2026 and lists only the SSE endpoint
- Rootly's Terms of Use bar crawling or scraping the Services by manual or automated means. This matters before any probe is run

## 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.

- Connect to `https://mcp.rootly.com/mcp-codemode` or add `?tool_profile=slim` to `/mcp`. The default endpoint loads about 218 tool definitions
- Connect Code Mode in place of the classic endpoint, not beside it. The docs say models skip `execute` when both are present
- Use a Global API key for `get_oncall_handoff_summary` and `get_oncall_shift_metrics`. Team and personal keys return partial results
- Refetch the tool list after a server release. Curated tools reject undeclared arguments, which failed calls from cached schemas in September 2026
- Treat a 404 from a generated tool as a possible plan limit or missing parent record, and read the hint in the response

## 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.