fix(elasticsearch): resolve Cloud IDs to the real host and correct the integration's output, timeout, and redirect defects

[waleedlatif1]

Summary

Fixes the Elastic Cloud endpoint bug, plus every defect a full /validate-integration pass turned up in this integration.

1. Cloud ID resolved to a host that does not exist (all 13 tools)

A Cloud ID is <deployment label>:<base64 of "parentDomain$esUuid$kibanaUuid"> and the reachable endpoint is https://<esUuid>.<parentDomain>. Every tool built https://<label>.<parentDomain> — the human-readable label, not the Elasticsearch UUID — so Elastic Cloud was unusable in every operation.

Replaced with parseCloudId in a new apps/sim/tools/elasticsearch/utils.ts, following Beats' libbeat/cloudid/cloudid.go decodeCloudID(): split at the last colon so a colon in the label cannot corrupt the payload; require ≥3 $-separated components; right-partition each component at its last colon for a per-service port, inheriting the parent domain's port and defaulting to 443.

Rejects the component characters Beats rejects (#@?/) plus \. This is the security-relevant part — an @ in the UUID component turns everything before it into URL userinfo, so https://<uuid>@evil.example.com would send Authorization: ApiKey … to an attacker-controlled origin. Two checks beyond Beats. First, both ports must be all digits — Beats validates neither, and uuid:80@evil.example.com survives its reject set by landing the @ in the port half. Second, : is rejected in a component name: extractPortFromName has already split at the last colon, so a colon surviving in the name half means the component carried two, and found.io:9243:5 would otherwise assemble https://<uuid>.found.io:9243:5 and fail as a bare TypeError: Invalid URL inside the transport. The port check does not cover that case — it only fires when the trailing half is non-numeric — so the two checks carry separate cases.

buildBaseUrl and buildAuthHeaders were duplicated across all 13 tool files, byte-identical apart from _bulk's NDJSON media type. Now shared, with _bulk passing its content type as an argument.

2. elasticsearch_get_index declared a phantom output

Declared index, but GET /{index} returns a map keyed by index name — {"logs-2024": {aliases, mappings, settings}}. There is no index key at any level, so the whole payload was unreferenceable from downstream blocks.

Now returns { indices: data, ...data } and declares indices. The raw per-index keys are spread alongside so references saved before indices existed keep resolving; an index legitimately named indices is spread last and wins. The error branch returns { indices: {} } so both branches satisfy the declared shape.

3. elasticsearch_cluster_health declared a param literally named timeout

tools/request-transport.ts reads params.timeout as the outbound HTTP deadline in milliseconds. On the block path this was inert — the mapper always suffixed s, and Number('30s') is NaN, which the transport discards. On the agent tool-calling path it was live: a model emitting timeout: 30 as a number skips the mapper's typeof === 'string' guard, and the generic handler merges raw inputs over the transform, so it arrived as a 30 ms client abort.

Renamed to clusterTimeout, mapped in tools.config.params (not tools.config.tool, which runs at serialization before variable resolution). The mapper also clears timeout explicitly, because the handler merges { ...inputs, ...transformedParams } and the raw input would otherwise still reach the transport. The subBlock id stays timeout, so saved workflow state is not orphaned — check-block-registry.ts subblock-ID stability passes.

Same mapper had a unit bug: it appended s to anything not already ending in s, so 1m became 1ms — a 1-millisecond server-side wait. It now only appends s to a bare integer.

4. Credentials survived a cross-origin redirect (all 13 tools)

Every tool sends Authorization, but nothing stripped it on a redirect off the configured origin. prepareToolRequest only populates redirectPolicy from the tool, and the stripping branch in lib/core/security/input-validation.server.ts is gated on that policy existing — so a redirect carried the API key or Basic credentials to the redirect target.

All 13 now declare { mode: 'legacy', sendCredentialsOnCrossOriginRedirect: false }. stripAuthOnRedirect is deliberately not used: it drops Authorization on every hop including same-origin, which would 401 a reverse proxy in front of Elasticsearch issuing a legitimate same-origin redirect. mode: 'legacy' preserves method and body replay — under 'standard', a 301/302 would rewrite POST to GET and break _search, _count and _bulk.

5. A cloud deployment fell back to a stale self-hosted host

buildBaseUrl branched on deploymentType === 'cloud' && cloudId, so a cloud invocation with no Cloud ID fell through to host. The block marks cloudId required and does not render host when cloud is selected, but a user who switches the dropdown keeps their previous host in saved state — hidden, not cleared. The cloud credential was sent to the cluster they used to point at.

Now branches on deployment type first and rejects a missing Cloud ID. An unrecognized deployment type is also rejected rather than falling through: deploymentType is required: true with no explicit visibility, which tools/params.ts resolves to user-or-llm, so a model supplies it on the agent path and a near miss like Cloud would take the self-hosted branch. A nullish value still means self-hosted — the dropdown's own default, and the shape of state saved before the field was touched.

6. list_indices dropped indices silently and could throw

item.index.startsWith('.') threw outright on a _cat row with no index column, and every system index was filtered out with no way to opt in. Now guarded, tolerant of a non-array body, and opt-in via an advanced includeSystemIndices dropdown. Default behavior is unchanged.

7. Nullable outputs missing optional: true

get_document._version and ._source (both absent on the 404 branch; _source also whenever _source_excludes strips it), delete_document._version, create_index.shards_acknowledged and .index, cluster_stats.status.

8. Block outputs did not cover every tool

list_indices.message and count._shards were returned but undeclared, so they were missing from the reference picker.

Backwards compatibility

Zero subBlock ids removed or renamed, zero required flips (one addition, a new optional param), zero visibility changes. check-block-registry.ts origin/staging passes the subblock-ID stability check.

Behavior only moves in one direction: a valid Cloud ID that produced an unreachable host now produces the reachable one; a malformed one threw before and throws now. get_index keeps its raw keys. list_indices keeps its default filter. The timeout subBlock keeps its id and its saved values.

Type of Change

• Bug fix

Testing

94 tests across utils.test.ts, cluster_health.test.ts, responses.test.ts and the existing search.test.ts.

Covers label-vs-UUID host resolution, colon-in-label, per-service and inherited ports, the #@?/\ reject set, both port-smuggle variants, <3 components, self-hosted trailing-slash and missing-host, a sweep asserting all 13 tools resolve the same cloud host, _bulk media type, prepareToolRequest leaving no client deadline while emitting timeout=30s, the 1m/500ms/2h unit cases, get_index declared-vs-actual output parity on both branches, list_indices filtering/opt-in/malformed rows, the redirect policy on all 13 tools, and the optional flags.

Every fix was verified red-first by reverting it and watching its tests fail: get_index 2, list_indices 3, redirect policy 10, optional flags 4, timeout mapper 6, Cloud ID reject set 3, cloud-host fallback 7.

bun run lint, bun run type-check (no Elasticsearch diagnostics), bun run check:audits (39 audits), and check-block-registry.ts origin/staging all pass. tool-metadata:generate and generate-docs artifacts regenerated and committed.

Checklist

• Code follows project style guidelines
• Self-reviewed my changes
• Tests added/updated and passing
• No new warnings introduced
• I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
docs	Ready	Preview	Aug 30, 2026 6:15pm

[greptile-apps]

Greptile Summary

The PR centralizes Elasticsearch connection/authentication handling and corrects Cloud ID resolution, redirect credential handling, cluster-health timeout mapping, and several response contracts.

• Resolves Elastic Cloud IDs to Elasticsearch service UUID hosts and rejects unsafe decoded components.
• Prevents credentials from crossing origins during redirects while preserving Elasticsearch request methods and bodies.
• Keeps saved cluster-health timeout fields compatible while separating the Elasticsearch wait from the HTTP client deadline.
• Aligns block, tool, generated, and documented outputs and adds system-index opt-in behavior.

Confidence Score: 5/5

The PR appears safe to merge.

No blocking failure remains.

Important Files Changed

Filename	Overview
apps/sim/tools/elasticsearch/utils.ts	Centralizes Cloud ID parsing, deployment URL selection, and authentication-header construction with explicit malformed-input checks.
apps/sim/blocks/blocks/elasticsearch.ts	Preserves the saved timeout sub-block ID while mapping it to the renamed tool parameter and adds system-index and output metadata.
apps/sim/tools/elasticsearch/cluster_health.ts	Separates Elasticsearch's server-side cluster wait from the transport timeout and applies safe redirect handling.
apps/sim/tools/elasticsearch/get_index.ts	Exposes the provider's index-keyed response through the declared indices output while retaining legacy top-level index keys.
apps/sim/tools/elasticsearch/list_indices.ts	Handles malformed response bodies and rows safely and adds an explicit option to include system indices.
apps/sim/tools/elasticsearch/utils.test.ts	Exercises Cloud ID host resolution, port inheritance, malformed components, authentication headers, and consistent tool behavior.
apps/sim/tools/elasticsearch/responses.test.ts	Verifies runtime response shapes and declared output metadata across the corrected Elasticsearch operations.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart LR
  A[Workflow or agent inputs] --> B{Deployment type}
  B -->|Cloud| C[Parse and validate Cloud ID]
  B -->|Self-hosted| D[Normalize configured host]
  C --> E[Build Elasticsearch endpoint]
  D --> E
  E --> F[Attach API key or Basic auth]
  F --> G[Execute request with legacy redirect semantics]
  G --> H{Redirect crosses origin?}
  H -->|Yes| I[Strip credentials]
  H -->|No| J[Retain credentials]
  I --> K[Transform structured output]
  J --> K

Loading

_{Reviews (6): Last reviewed commit: "Merge remote-tracking branch 'origin/sta..." | Re-trigger Greptile}

[cubic-dev-ai]

All reported issues were addressed across 16 files

<sub>Reply with feedback, questions, or to request a fix.

Fix all with cubic | Re-trigger cubic</sub>

[waleedlatif1]

@greptile

[waleedlatif1]

@cubic review

[cubic-dev-ai]

@cubic review

@waleedlatif1 I have started the AI code review. It will take a few minutes to complete.

[cubic-dev-ai]

All reported issues were addressed across 16 files

<sub>Reply with feedback, questions, or to request a fix.

Fix all with cubic | Re-trigger cubic</sub>

[waleedlatif1]

@greptile

[waleedlatif1]

@cubic review

[cubic-dev-ai]

@cubic review

@waleedlatif1 I have started the AI code review. It will take a few minutes to complete.

[cubic-dev-ai]

All reported issues were addressed across 22 files

<sub>Requires human review: Auto-approval blocked because this review re-detected 1 unresolved issue already reported by Cubic.

Re-trigger cubic</sub>

[waleedlatif1]

@greptile

[waleedlatif1]

@cubic review

[cubic-dev-ai]

@cubic review

@waleedlatif1 I have started the AI code review. It will take a few minutes to complete.

[cubic-dev-ai]

All reported issues were addressed across 22 files

<sub>Reply with feedback, questions, or to request a fix.

Fix all with cubic | Re-trigger cubic</sub>

[waleedlatif1]

@greptile

[waleedlatif1]

@cubic review

[cubic-dev-ai]

@cubic review

@waleedlatif1 I have started the AI code review. It will take a few minutes to complete.

[cubic-dev-ai]

No issues found across 22 files

Confidence score: 5/5

• Automated review surfaced no issues in the provided summaries.
• No files require special attention.

<sub>You've manually re-run cubic several times on this PR. Each manual re-review checks the full PR again and counts toward your usage quota. To preserve your usage limits, we recommend letting cubic automatically review new commits.

Re-trigger cubic</sub>

[waleedlatif1]

@greptile

[waleedlatif1]

@cubic review

[cubic-dev-ai]

@cubic review

@waleedlatif1 I have started the AI code review. It will take a few minutes to complete.

[cubic-dev-ai]

No issues found across 22 files

Confidence score: 5/5

• Automated review surfaced no issues in the provided summaries.
• No files require special attention.

<sub>You've manually re-run cubic several times on this PR. Each manual re-review checks the full PR again and counts toward your usage quota. To preserve your usage limits, we recommend letting cubic automatically review new commits.

Re-trigger cubic</sub>