Elasticsearch

# Fix list: Elasticsearch

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/elasticsearch, the October 2026 research run, assessed 8 October 2026. Grade B, 67.7 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 Elasticsearch: 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, 35 out of 100, up to 8.1 more on the total

Why it scored 35: No x402, MPP or L402 in the docs index, the OpenAPI file or the pricing page (0). Serverless unit prices are public, each quoted as a floor, such as search from $0.09 a VCU-hour and storage from $0.047 a GB a month (15 of 20). 14-day trial with no credit card (20). A person signs up in a browser and creates the first key (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, 62 out of 100, up to 7.6 more on the total

Why it scored 62: Graded as a hosted service on Elastic Cloud Serverless. Statuspage at status.elastic.co with 638 components by region (20). Between 10 July and 8 October 2026 it lists more than 20 notices marked major. Most concern provisioning, metrics or the console, but three are regional service incidents, GCP us-central1 for 4 hours 47 minutes on 1 September, high latency in GCP asia-south1 for 3 hours on 6 August, and Serverless projects in GCP us-central1 delayed or unavailable on 8 October and still open when read. Elasticsearch 9.5.0 also returned incorrect results for some queries from 5 to 20 August (5 of 30). No request rate limit with numbers found, only project limits (5 of 15). The API conventions page says to retry 429 after a delay with exponential backoff, with no `Retry-After` header documented, and `_create` with an explicit ID makes a retried write safe (12 of 15). Serverless SLA of 99.95 per cent at Platinum or Enterprise level (10). The REST API and the MCP server are generally available on Serverless (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.

## 3. Security & auth, 72 out of 100, up to 4.9 more on the total

Why it scored 72: API keys take role descriptors that limit cluster, index and Kibana privileges and carry an expiry, and the MCP server accepts OAuth 2.1 on Serverless with per-user connections that can be revoked (30). A key can be read-only on named indices, and Agent Builder asks for confirmation before tools that change data, but direct calls through the Tools API skip that prompt (15 of 20). Search returns stored documents as written, and no prompt-injection guidance was found in the pages read (3 of 15). Agent traces can be collected into `traces-agent_builder.otel-*`. Audit logging on Serverless is unchecked (6 of 15). No security.txt (404). HackerOne bug bounty, CVE Numbering Authority, public advisories, SOC 2 and ISO 27001 (18 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.

## 4. Agent ergonomics, 77 out of 100, up to 3.7 more on the total

Why it scored 77: The API sizes responses with `size` and `_source` includes and excludes. The MCP server exposes the Agent Builder tool list, 88 built-in tools in the reference with 21 in `platform.core`, and a key's privileges decide which ones work (15 of 25). `from` and `size`, `search_after`, filters and sorting (20). Errors come back as JSON with a type, reason and status, and the 429 and rejected-request pages say how to recover (15 of 20). `_create` and explicit document IDs make writes safe to retry. MCP annotations unchecked (12 of 20). Official clients for JavaScript and Python at 9.5.1, among others (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. Transparency & trust, 71 out of 100, up to 2.5 more on the total

Made of editorial 70, provenance 71.

Why it scored 71: Source is under AGPL-3.0, SSPL-1.0 or Elastic Licence 2.0, with `x-pack` under Elastic Licence 2.0 only, and Serverless itself is closed (22 of 30). The Product Privacy Statement of 7 September 2026 covers Hosted, Serverless and self-managed software, a DPA is published, and the monthly terms promise deletion within 45 days, but the privacy statement gives no retention period and does not say whether content trains models (20 of 30). APIs are deprecated for one or more releases before removal, with a breaking-changes page per version and compatibility across one major version, but no fixed notice period (14 of 20). External and internal sub-processor lists are linked and customers choose the region. We did not read the lists (14 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: elastic.co, no registry record we could read (0 of 15)
- Terms of service: published, but our reader couldn't read it (7 of 10)
- Privacy policy: read, states 7 of the 8 things a reader expects (9.3 of 10)
- security.txt: not found (0 of 10)

## 6. Schema & documentation, 88 out of 100, up to 2 more on the total

Why it scored 88: OpenAPI 3.0.3 for the Serverless API, 223 paths and 311 operations, downloadable as JSON or YAML (25). llms.txt and a Markdown twin of each docs page (10). 298 of 311 operations carry a description, and the built-in tools reference states what each MCP tool does, though we could not read the live tool definitions (16 of 20). Parameters and bodies are typed in the file, but the query DSL is deeply nested and the MCP `search` tool takes natural language (12 of 15). 292 operations carry examples. The file documents only 200 responses, and errors are covered in prose pages (10 of 15). Release notes per version, a breaking-changes page and compatibility headers (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. Maintenance & community, 83 out of 100, up to 1.5 more on the total

Why it scored 83: Elasticsearch 9.5.5 was tagged on 30 September 2026 and announced on 6 October (30). 9.5.3, 9.5.4, 9.5.5 and three 9.4 patch releases fall inside 90 days (20). Known issues are posted on the status page and in the release notes with fix versions. GitHub issue handling is unchecked because the API refused us (12 of 25). JavaScript and Python clients at 9.5.1, matching the server line (15). Build and CI configuration is in the repository. The 9.5 line needed three patch releases for correctness and replication defects (6 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.

## Deductions

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

- Sixteen Elasticsearch security advisories were posted between 25 September and 6 October 2026. The one we read, ESA-2026-199 (CVE-2026-103009, CVSS 7.1), is an authorisation bypass in cross-cluster search on self-managed clusters, fixed in 8.19.23, 9.4.8 and 9.5.5, and Elastic says Serverless is not affected. Fixed and disclosed, so a small deduction (https://discuss.elastic.co/c/announcements/security-announcements/31).

## 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 star count and issue responsiveness, because the GitHub API refused us for its rate limit
- unchecked: the MCP server's live tool definitions, its transport, and whether tools carry readOnlyHint or destructiveHint
- unchecked: audit logging on Serverless. The docs page we tried returned 404
- unchecked: the contents of the external and internal sub-processor lists linked from elastic.co/trust/privacy
- unchecked: the registration date of elastic.co. No RDAP service answered for the domain
- The MCP options page, as our reader summarised it, labels the Agent Builder MCP server preview, while the server's own page says generally available on Serverless and since 9.3. We followed the server's page
- Whether the 9.5.0 query defect reached Serverless projects. The status notice names versions, not deployment types
- The lead called the MCP server a preview on Serverless. Elastic's page for it says generally available
- The Cloud Monthly Terms and the Cloud Service Subscription Agreement are published as PDFs behind cover pages, which our policy reader does not fetch
- www.elastic.co/.well-known/security.txt returned 404

## Weaknesses

- More than 20 status notices marked major between 10 July and 8 October 2026, three of them regional service incidents
- Elasticsearch 9.5.0 (29 July 2026) returned incorrect results for some `must_not` queries with no error, fixed in 9.5.1 and 9.5.2
- No request rate limit with numbers was found in the reviewed documentation
- The 99.95 per cent Serverless SLA applies only at Platinum or Enterprise subscription level
- The monthly terms forbid use of the cloud service for benchmarking, competitive or comparative purposes
- Search compute on Serverless does not scale to zero, so an idle project is still billed a baseline

## 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.

- Send `Authorization: ApiKey <key>` to the project's Elasticsearch URL. The MCP server is on the Kibana URL at `/api/agent_builder/mcp`
- Create the MCP key with `feature_agentBuilder.read` and only the index patterns needed, or the endpoint answers 403
- On 429, wait and retry with exponential backoff. Do not set short client timeouts, because a retried request joins the back of the queue
- Page past 10,000 hits with `search_after`, not `from` and `size`
- Tool calls made directly through the Tools API skip Agent Builder's confirmation prompts, so restrict the key instead

## 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.