# Fix list: Grafana MCP Server From Anchor Terminal's listing at https://www.anchorterminal.com/tools/grafana-mcp-server, the October 2026 research run, assessed 8 October 2026. Grade BB, 74.5 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 Grafana 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. Security & auth, 65 out of 100, up to 6.1 more on the total Why it scored 65: The server holds a Grafana service account token, which Grafana RBAC can scope by action and resource and which can be revoked. `GRAFANA_SERVICE_ACCOUNT_TOKEN_FILE` is re-read on each request for rotation. Username and password and a deprecated API key also work. The open-source server has no OAuth, and no secret travels in a URL (25 of 30). `--disable-write`, `--disable-query`, per-category flags, Loki enforced matchers and a per-tool RBAC table support least privilege. Writes are on by default, the default set includes `grafana_api_request` for any API path, and nothing asks for confirmation (14 of 20). Log lines, dashboards and annotations reach the model as stored. No injection guidance was found in the README or docs, and issue #680 on the subject has been open since 24 March 2026 (2 of 15). OpenTelemetry traces, metrics and logs per tool call and slow-request logging, with arguments left out of spans unless `--include-args-in-spans` is set (10 of 15). Grafana Labs runs a disclosure programme on Intigriti without bounties, lists ISO 27001 and SOC 2 Type 2, and publishes advisories for its products. This repository has no SECURITY.md and no published advisories, though its changelog records security fixes, and grafana.com has no security.txt (14 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. ## 2. Payments & pricing, 60 out of 100, up to 5 more on the total Why it scored 60: Read with the self-hosted rule. The server is free under Apache-2.0 and needs no signup of its own, so 20 + 20 + 20. No x402, MPP or L402 (0). It needs a Grafana instance, which can be the free open-source Grafana. Grafana Cloud prices were not checked. The separate hosted Grafana Cloud MCP server counts towards Grafana Assistant usage and is not what this listing grades. 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. Agent ergonomics, 74 out of 100, up to 4.2 more on the total Why it scored 74: 79 tools load by default in 2.0.2, about 107 KB of definitions as JSON, and 119 with every category on. That is 5 of 25 by count, plus 9 for `--enabled-tools`, per-category `--disable-*` flags, 13 opt-in categories and `--disable-write`, which leaves 62 (14). `limit` and `page` on searches, `get_dashboard_summary`, JSONPath reads through `get_dashboard_property`, a jq filter on `grafana_api_request`, and a 100-line default on `query_loki_logs` with an optional byte and time-range guardrail (18 of 20). Unknown argument keys are rejected with the valid ones listed, empty results carry likely causes and next steps, and TraceQL errors return correction hints. No documented error list (16 of 20). Every default tool sets `readOnlyHint`, `destructiveHint` and `idempotentHint` (60 read-only, 9 marked destructive). No idempotency keys on creates (17 of 20). 85 of 383 parameters are required and most have defaults. The Go module is the only library, with a Python wrapper for install (9 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. ## 4. Schema & documentation, 81 out of 100, up to 3.1 more on the total Why it scored 81: All 79 default tools declare typed JSON Schema inputs with `additionalProperties` false, read from `tools/list` on 2.0.2. No tool declares an output schema, which issue #1278 asks for (23 of 25). grafana.com serves llms.txt and a Markdown form of every docs page, the MCP pages included (10). Descriptions state purpose and inputs, with a median of 240 characters. Version 2.0.2 removed model directives and references to other tools from descriptions to meet connector directory rules, so guidance on when to use a tool is thinner than it was (14 of 20). 375 of 383 parameters carry a description, but only 14 are enums. Values such as `queryType` ('range' or 'instant') and incident `status` are named in prose, and 16 parameters are free-form objects (9 of 15). The docs have client configuration examples, task guides and a troubleshooting page, and empty results return hints. No catalogue of error responses was found (10 of 15). Semantic versions, tagged releases and a dated changelog (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. ## 5. Reliability, 91 out of 100, up to 1.8 more on the total Why it scored 91: Scored as local software, since the owner runs it over stdio, SSE or streamable HTTP. Official PyPI package `mcp-grafana` (a wheel that carries the Go binary, Python 3.10 or later), a Docker image, release binaries and a Helm chart, with Grafana 9.0 or later stated as the requirement (20). Unit, integration and end-to-end workflows all concluded in success on main on 8 October 2026 (25). 54 open issues, none labelled bug, and 38 closed in the 30 days before the check. Three bug reports remain open from July and August (#992 `run_panel_query` 422, #1005 `grafana_api_request` 500 on plugin resources, #1049 a dropped table row in `get_panel_image`), each with replies (19 of 25). The changelog follows Keep a Changelog and semantic versioning and labels breaking changes, and 1.6.2 announced that 2.0 came next. Patch release 1.4.1 carried a breaking parameter change on the Sift tools, and 2.0.0 followed 1.0.0 by two months (12 of 15). Version 2.0.2 (15). 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. ## 6. Transparency & trust, 79 out of 100, up to 1.8 more on the total Made of editorial 81, provenance 76. Why it scored 79: Apache-2.0 (30). Local software that talks to the configured Grafana. A docs page lists every field in the usage report, the destination, and the region and network name the receiving service adds, and the Grafana Labs privacy policy (last updated 10 July 2026) covers usage statistics from its open-source software. No retention period for the reports was found (24 of 30). No deprecation policy. Notices come without dates, such as `GRAFANA_API_KEY` to be removed 'in a future version', and the Sift tools were marked deprecated before 2.0.0 removed them (10 of 20). Usage statistics are documented with three opt-outs and a `log` mode that prints the report without sending it. They arrived off by default in 1.5.0 on 17 September 2026 and were switched on by default in 2.0.0, noted in the changelog (17 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): - Status page: not found (0 of 10) - security.txt: not found (0 of 10) ## 7. Maintenance & community, 95 out of 100, up to 0.4 more on the total Why it scored 95: v2.0.2 tagged on 7 October 2026, one day before the check (30). 17 tagged releases from 0.17.2 on 13 July to 2.0.2 (20). 38 issues closed in the 30 days before the check against 54 open and 22 open pull requests, with commits on main most days. Several feature requests from March and April have no reply (21 of 25). Listed in the official MCP registry as `io.github.grafana/mcp-grafana`, a GitHub-verified namespace, with v2.0.1 of 6 October marked latest in the page we read (14 of 15). Renovate, a Docker image scan workflow and, from 8 October, cosign signing and SBOMs for releases (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. ## Deductions Each comes off the total. A fixed and documented problem counts for less at the next check. - -3: before 0.17.2 (13 July 2026) a URL in the then undocumented `X-Grafana-URL` request header caused the server's environment service account token, API key, basic auth or extra headers to be sent to a host the caller chose, on the HTTP transports. Fixed and described in the changelog, whose entry links a placeholder pull request number, and no advisory was published, so a reduced deduction (https://github.com/grafana/mcp-grafana/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. - The lead did not mention the hosted Grafana Cloud MCP server at mcp.grafana.com, a separate product with OAuth 2.1. This listing grades the open-source server only - unchecked: Grafana Cloud pricing and its free tier, which matter only when the server points at Grafana Cloud - unchecked: the Helm chart in grafana/helm-charts and the Docker Hub image page, which rest on the README - unchecked: whether the MCP registry lists 2.0.2, since the page we read ended at v2.0.1 marked latest - unchecked: the Intigriti programme page and its scope, and whether mcp-grafana is in scope - The 0.17.2 changelog entry links pull request #XXX. The commit on 13 July 2026 names #995 - No retention period for usage reports was found in the usage statistics docs or the privacy policy ## Weaknesses - The default set is 79 tools and about 107 KB of definitions, and 119 tools with every category on - Write tools and `grafana_api_request`, which can call any Grafana API path, are enabled unless the operator turns them off - Anonymous usage statistics went from off by default in 1.5.0 (17 September 2026) to on by default in 2.0.0 - A credential-forwarding flaw fixed in 0.17.2 on 13 July 2026 has no published advisory, and the repository has no SECURITY.md - No prompt-injection guidance for log lines and dashboard content. Issue #680 on the subject has been open since 24 March 2026 ## 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. - Start with `--disable-write` unless the task needs writes, and add `--disable-api` to remove `grafana_api_request` - Pass `--enabled-tools` with only the categories needed. The default list loads 79 tools - Use `get_dashboard_summary` or `get_dashboard_property` with a JSONPath before `get_dashboard_by_uid`, which returns the whole dashboard JSON - Set `GRAFANA_USAGE_STATS=disabled` or `DO_NOT_TRACK=1` to stop the four-hourly usage report to stats.grafana.org - Use streamable HTTP, not SSE, when each caller needs its own token or organisation, and set `--server-auth-token` on any non-loopback address ## 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.