Docs fix — style: path /explore-analyze/ai-features — 75 pages

Generated by `gh-aw-docs-style-sweep` for `elastic/docs-content` on 2026-W25.

Path /explore-analyze/ai-features · 75 total in scope · subtree corpus 75 pages.

## Findings (29)

```yaml
- file: explore-analyze/ai-features/agent-builder/agent-builder-dashboards-and-visualizations.md
line: 60
category: accessibility
severity: medium
evidence: "'on the left' in alt text — Elastic.DirectionalLanguage rule"
suggested_fix: |
Replace alt text directional phrasing: "Full-screen chat with a conversation panel and a dashboard canvas panel showing metrics, charts, and trend panels"

- file: explore-analyze/ai-features/agent-builder/agent-builder-dashboards-and-visualizations.md
line: 60
category: accessibility
severity: medium
evidence: "'on the right' in alt text — Elastic.DirectionalLanguage rule"
suggested_fix: |
Remove directional reference; describe content instead of position (see finding above)

- file: explore-analyze/ai-features/agent-builder/connectors.md
line: 41
category: ui-writing
severity: medium
evidence: "'**Manage components** > **Connectors**' — Elastic.MenuArrowsBold rule; '>' should be '→'"
suggested_fix: |
select **Manage components → Connectors**

- file: explore-analyze/ai-features/agent-builder/custom-skills.md
line: 41
category: ui-writing
severity: medium
evidence: "'**Manage components > Skills**' — Elastic.MenuArrowsBold rule; '>' should be '→'"
suggested_fix: |
Go to **Manage components → Skills**, then select **Create a skill**.

- file: explore-analyze/ai-features/agent-builder/mcp-server.md
line: 42
category: word-choice
severity: medium
evidence: "'etc.' — Elastic.Latinisms rule; Latin abbreviation, use plain English"
suggested_fix: |
Most MCP clients (such as Claude Desktop, Cursor, and VS Code) have similar configuration patterns.

- file: explore-analyze/ai-features/agent-builder/plugins.md
line: 29
category: ui-writing
severity: medium
evidence: "'**Manage components** > **Plugins**' — Elastic.MenuArrowsBold rule; '>' should be '→'"
suggested_fix: |
select **Manage components → Plugins**

- file: explore-analyze/ai-features/agent-builder/tools/custom-tools.md
line: 110
category: accessibility
severity: medium
evidence: "'in the left sidebar' — Elastic.DirectionalLanguage rule"
suggested_fix: |
Select the agent from the agent selector in the sidebar.

- file: explore-analyze/ai-features/agent-builder/tools/workflow-tools.md
line: 46
category: ui-writing
severity: medium
evidence: "'Agents > More > View all tools > New tool' — Elastic.MenuArrows rule; '>' should be '→'"
suggested_fix: |
Go to **Agents → More → View all tools → New tool**.

- file: explore-analyze/ai-features/agent-builder/tools/workflow-tools.md
line: 95
category: accessibility
severity: medium
evidence: "'in the left sidebar' — Elastic.DirectionalLanguage rule"
suggested_fix: |
Select the agent from the agent selector in the sidebar.

- file: explore-analyze/ai-features/agent-builder/tools/workflow-tools.md
line: 105
category: ui-writing
severity: medium
evidence: "'More > Edit Agent > Tools' — Elastic.MenuArrows rule; '>' should be '→'"
suggested_fix: |
Select **More → Edit Agent → Tools**.

- file: explore-analyze/ai-features/ai-chat-experiences/ai-assistant.md
line: 48
category: grammar
severity: medium
evidence: "\"Provide comments explaining the query you just gave\". — Elastic.QuotesPunctuation rule; period outside closing quotation mark"
suggested_fix: |
"Provide comments explaining the query you gave."

- file: explore-analyze/ai-features/ai-chat-experiences/ai-assistant.md
line: 48
category: word-choice
severity: medium
evidence: "'just' — Elastic.DontUse rule; avoid filler word 'just'"
suggested_fix: |
the query you gave

- file: explore-analyze/ai-features/llm-guides/connect-to-azure-openai.md
line: 57
category: grammar
severity: medium
evidence: "\"Auto-update to default\". — Elastic.QuotesPunctuation rule; period outside closing quotation mark"
suggested_fix: |
Set the model version to "Auto-update to default."

- file: explore-analyze/ai-features/llm-guides/connect-to-vLLM.md
line: 15
category: accessibility
severity: medium
evidence: "'The steps below' — Elastic.DirectionalLanguage rule"
suggested_fix: |
The following steps show one example configuration, but you can use any model supported by vLLM, including private and gated models on Hugging Face.

- file: explore-analyze/ai-features/agent-builder/agent-builder-agents.md
line: 46
category: word-choice
severity: low
evidence: "'clone it and create a custom agent' — Elastic.Clone rule; 'clone' used for duplicating an agent, not a Git repo"
suggested_fix: |
you can duplicate it and create a custom agent

- file: explore-analyze/ai-features/agent-builder/agent-builder-agents.md
line: 81
category: word-choice
severity: low
evidence: "'clone' in UI action list — Elastic.Clone rule; agent copy action uses 'clone' where 'duplicate' is clearer in prose"
suggested_fix: |
You can **chat**, **edit**, **duplicate**, or **delete** custom agents.

- file: explore-analyze/ai-features/agent-builder/builtin-agents-reference.md
line: 23
category: word-choice
severity: low
evidence: "'clone it as a starting point for a new agent' — Elastic.Clone rule"
suggested_fix: |
duplicate it as a starting point for a new agent

- file: explore-analyze/ai-features/agent-builder/builtin-agents-reference.md
line: 28
category: word-choice
severity: low
evidence: "'To customize it, clone it' — Elastic.Clone rule"
suggested_fix: |
To customize it, duplicate it and create a custom agent.

- file: explore-analyze/ai-features/agent-builder/custom-agents.md
line: 180
category: word-choice
severity: low
evidence: "'can only be chatted with or cloned' — Elastic.Clone rule"
suggested_fix: |
Other built-in agents can only be chatted with or duplicated, not edited or deleted.

- file: explore-analyze/ai-features/agent-builder/get-started.md
line: 155
category: word-choice
severity: low
evidence: "'take action' — Elastic.Wordiness rule; verbose phrase"
suggested_fix: |
or act

- file: explore-analyze/ai-features/agent-builder/limitations-known-issues.md
line: 55
category: word-choice
severity: low
evidence: "'may not be compatible' — Elastic.WordChoice rule; prefer 'might' over 'may' for possibility"
suggested_fix: |
might not be compatible

- file: explore-analyze/ai-features/agent-builder/limitations-known-issues.md
line: 67
category: word-choice
severity: low
evidence: "'may generate invalid ES|QL' — Elastic.WordChoice rule"
suggested_fix: |
might generate invalid ES|QL

- file: explore-analyze/ai-features/agent-builder/limitations-known-issues.md
line: 69
category: word-choice
severity: low
evidence: "'may generate invalid ES|QL' — Elastic.WordChoice rule"
suggested_fix: |
might generate invalid ES|QL

- file: explore-analyze/ai-features/agent-builder/models.md
line: 174
category: word-choice
severity: low
evidence: "'models may produce errors' — Elastic.WordChoice rule; prefer 'might'"
suggested_fix: |
Smaller or less capable models might produce errors like:

- file: explore-analyze/ai-features/agent-builder/skills.md
line: 25
category: word-choice
severity: low
evidence: "'making expertise easy to share' — Elastic.WordChoice rule; 'easy' is a condescending simplicity claim"
suggested_fix: |
making expertise reusable across your team

- file: explore-analyze/ai-features/agent-builder/tools/esql-tools.md
line: 33
category: word-choice
severity: low
evidence: "'Execute pre-defined ES|QL queries' — Elastic.WordChoice rule; prefer 'run' over 'execute'"
suggested_fix: |
Run pre-defined ES|QL queries with dynamic parameters

- file: explore-analyze/ai-features/agent-builder/tools/esql-tools.md
line: 89
category: grammar
severity: low
evidence: "'don't provide values. Without' — Elastic.Negations rule; stacked negations reduce clarity"
suggested_fix: |
they are strongly recommended to prevent query syntax errors. Without defaults, unspecified optional parameters will be `null`, which can cause queries to fail.

- file: explore-analyze/ai-features/elastic-agent-builder.md
line: 59
category: word-choice
severity: low
evidence: "'take action' — Elastic.Wordiness rule; verbose phrase"
suggested_fix: |
act, or support workflows

- file: explore-analyze/ai-features/llm-guides/connect-to-lmstudio-observability.md
line: 54
category: word-choice
severity: low
evidence: "'in order to work with Elastic' — Elastic.Wordiness rule"
suggested_fix: |
Your chosen model must include `instruct` in its name to work with Elastic.
```

## Done when
- All listed lines pass the Elastic style guide rules cited.
- A PR addressing this issue is merged.

## Notes
- Skipped `Elastic.MappedPages` warnings (front matter metadata, not a style issue for writers). Skipped `Elastic.Latinisms` findings for "VS Code" (proper product name). Skipped `Elastic.Clone` findings in `agent-skills.md` (all refer to Git repository cloning, which is the approved use). Skipped `Elastic.EndPuntuaction` findings (lines were prose sentences, not headings — likely Vale false positives). Skipped `execute` findings where the match was inside a snippet `:::{include}` filename or a code identifier.

> Generated by [Docs style sweep agent](https://github.com/elastic/docs-content/actions/runs/27846532575) · 525.1 AIC · ⌖ 13.9 AIC · ⊞ 26.7K · [◷](https://github.com/search?q=repo%3Aelastic%2Fdocs-content+is%3Aissue+%22gh-aw-workflow-call-id%3A+elastic%2Fdocs-content%2Fgh-aw-docs-style-sweep%22&type=issues)

Start with the listed Markdown files and line numbers, reviewing each cited Elastic style rule and its suggested fix. Check the affected pages with the documentation style tooling used by gh-aw-docs-style-sweep. Done means all 29 findings pass their cited rules and a pull request addressing the issue is merged.

Written by the indexing model from the issue text.