Datadog MCP Server by Datadog

# Fix list: Datadog MCP Server

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/datadog-mcp, the October 2026 research run, assessed 1 October 2026. Grade C, 56.7 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 Datadog 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. Payments & pricing, 20 out of 100, up to 10 more on the total

Why it scored 20: No x402, MPP or L402 (0). Datadog plan prices are public and the MCP quota of 100,000 calls a month per organisation is published, but there's no MCP price or per-call rate (10). A free Datadog plan exists per the 30 September check, and nothing we read says whether it includes MCP, so half (10). A person signs up and authorises OAuth or creates keys (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.

## 2. Reliability, 53 out of 100, up to 9.4 more on the total

Why it scored 53: Scored as a hosted MCP server. status.datadoghq.com is an Atlassian Statuspage with more than 40 US1 components and a separate EU page, but no MCP component (15). We could read only the front page. The JSON feed was refused by our fetch proxy's rate limit and the history page didn't render. It shows three incidents in the ten days before this check, delayed monitor notifications on 21 September, On-Call pages that couldn't be acknowledged or resolved on 25 September and delayed events on 30 September, none touching the query APIs the MCP server calls (10). The changelog publishes a burst limit of 50 calls per 10 seconds and a monthly quota of 100,000 tool calls per organisation, doubled from 50,000 on 28 August (15). Read-only software-delivery calls are retried on transient failures and one tool explains rate limiting, but we found no documented Retry-After or backoff guidance for MCP (5). No SLA found (0). GA since 9 March 2026 and on the stable /v1 URL since 20 July, with some toolsets still experimental (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.

## 3. Schema & documentation, 60 out of 100, up to 6.5 more on the total

Why it scored 60: Tool schemas are only visible through tools/list with an account. The changelog shows typed parameters, for example `limit` 1 to 1,000 and percentile enums, so we gave partial credit for a contract we couldn't read (15). We couldn't check docs.datadoghq.com/llms.txt because our fetch proxy refused it, and the listing doesn't record one (0). Changelog entries show descriptions being corrected, such as `search_pr_insights` now explaining that `expected` means pending (10). Ranges, cursors, enums and attribute selectors per the changelog (11). Errors are being made actionable, with reversed time windows rejected up front and an explicit `result_too_large` failure (9). A dated public changelog with 102 entries since 9 March 2026 and a versioned /v1 URL (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.

## 4. Security & auth, 76 out of 100, up to 4.2 more on the total

Why it scored 76: OAuth 2.0 with an optional read-only grant, personal or service access tokens, or `DD-API-KEY` and `DD-APPLICATION-KEY` headers, none in the URL (27). `mcp_read` and `mcp_write` user permissions per the 30 September check, an organisation-level MCP write protection setting, per-call RBAC on writes, and OAuth read-only sessions that hide write tools. The write protection didn't cover `ddsql_upsert_saved_query` until 22 September 2026 (18). Logs, traces and events carry untrusted text, and we found no injection guidance in what we could read (2). Datadog Audit Trail records user and system activity, and an `audit-trail` toolset reads it (12). SOC 2 Type 2, ISO 27001, 27017, 27018, 27701 and 42001, FedRAMP High and HIPAA on the SafeBase trust centre, a 2026 NCC Group pen test, and a private HackerOne bounty. No security.txt, which returns 404 (17).

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.

## 5. Agent ergonomics, 78 out of 100, up to 3.6 more on the total

Why it scored 78: More than 30 toolsets chosen with the `toolsets` parameter, `omit_tools` per the 30 September check, `whoami` and `toolsets` resources, and a core set trimmed by about 15 per cent in March. An `execute_code` toolset runs sandboxed TypeScript against the APIs instead of one tool per call. We couldn't count the default tools (20). `limit`, cursors, `custom_attributes`, token budgets and flattening on aggregates, and `exclude_fields` on Kubernetes manifests (19). The changelog shows a steady effort on actionable errors (15). OAuth read-only grants omit write tools, an org setting blocks MCP writes, and read-only calls retry automatically. We couldn't confirm tool annotations (12). Official Datadog API clients in several languages and Python agent examples in the repository (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.

## 6. Maintenance & community, 68 out of 100, up to 2.8 more on the total

Why it scored 68: Changelog entry on 25 September 2026 (30). 42 dated entries since 3 July (20). Closed service with a weekly public changelog and Datadog support, and the examples repository merges its own pull requests within days (13). Not in the official MCP registry. Every Datadog entry there is a third-party server (0). Official API clients are current, but the examples repository has no CI (5).

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.

## 7. Transparency & trust, 68 out of 100, up to 2.8 more on the total

Made of editorial 56, provenance 80.

Why it scored 68: The hosted server is closed with published terms, and the examples repository is MIT (18). Privacy policy, DPA and a trust centre exist, but we found nothing MCP-specific on what the server logs or keeps (18). The changelog labels breaking changes and deprecations, for example `get_datadog_spreadsheet_table_data` deprecated on 20 July, but removals ship the day they're announced, and nothing says whether the old /api/unstable/ URL still works (8). Datadog sites per region are documented. The subprocessor list sits behind the trust centre's request form (12).

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):

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

## Deductions

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

- -3: breaking changes shipped the day they were announced, with no notice period. Detection-rule tools replaced and removed on 2026-06-17, `service` and `family` removed from `explore_profiling_call_graph` on 2026-06-26, `start_at` removed from `search_datadog_spans` on 2026-08-20 and the `traces` extension removed from `execute_code` on 2026-09-24. Each is labelled in the changelog, so we deduct at the low end (https://github.com/datadog-labs/mcp-server/blob/main/CHANGELOG.md).
- -1: until 2026-09-22, `ddsql_upsert_saved_query` created or updated saved queries even when an organisation had disabled MCP writes. Fixed and disclosed in the changelog, so a small deduction (https://github.com/datadog-labs/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: the MCP docs pages (docs.datadoghq.com/bits_ai/mcp_server/ and its setup page) and docs.datadoghq.com/llms.txt, which our fetch proxy refused for rate limiting and told us not to retry. Setup facts such as `mcp_read`, `mcp_write`, `omit_tools` and the preview toolsets rest on the listing's 30 September check
- unchecked: the 90-day incident history, since the status feed was refused and the history page didn't render
- unchecked: the default tool count and whether tools carry readOnlyHint and destructiveHint
- unchecked: whether the old /api/unstable/mcp-server/mcp URL still answers
- unchecked: whether the free Datadog plan includes MCP access

## Weaknesses

- Tools and parameters are removed the day they're announced, with no notice period
- Not in the official MCP registry and no MCP component on the status page
- Tool schemas and descriptions aren't readable without an account
- No security.txt, and the write-protection setting missed one tool until 22 September 2026
- Whether the free plan includes MCP isn't stated

## 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.

- Use https://mcp.datadoghq.com/v1/mcp. The /api/unstable/ path in older configs is no longer the documented one
- Set `toolsets` to the two or three you need, for example core plus logs, and read the `toolsets` resource first
- Pass `limit` and `custom_attributes` on span searches, and page with `cursor`. `start_at` is gone
- Expect a permission error when the org blocks MCP writes or the OAuth grant is read-only. That's the boundary working
- Budget calls. The org shares 100,000 a month and 50 per 10 seconds

## What the review panel asked for

- notice period before removals
- dated sunset for /api/unstable/
- publish tool definitions
- notice before removals

## 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.