docs-page-opening-optimizer
Skill72 starsChanged 5 months agoApache-2.0View on GitHub
Optimize the opening of an Elastic documentation page — H1 title, opening paragraph, and requirements section — following doc type conventions. Use when writing or improving page intros, optimizing titles for discoverability, adding requirements sections, or when the user asks to improve the first lines of a doc page.
In this repository26 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- create-skillSkill
- docs-applies-to-taggingSkill
- docs-content-type-checkerSkill
- docs-draft-feature-docsSkill
- docs-frontmatter-descriptionSkill
- docs-redirectsSkill
- docs-syntax-helpSkill
- docs-fix-changelogSkill
- docs-kibana-release-notesSkill
- docs-review-changelogSkill
- docs-serverless-changelogSkill
- docs-draft-workflow-docsSkill
- docs-kibana-settingsSkill
- docs-lens-chart-pageSkill
- docs-lens-chart-settingsSkill
- docs-check-contradictionsSkill
- docs-check-kibana-oasSkill
- docs-check-styleSkill
- docs-flag-jargon-skillSkill
- docs-frontmatter-auditSkill
- docs-review-prSkill
- docs-skill-reviewSkill
- docs-validate-code-samplesSkill
- docs-file-issueSkill
- docs-stack-releaseSkill
The file
We split this skill into the rules it gives an agent
11 rules. About 0 are common, 7 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈7
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 38
Read the file and identify its type.
- Very specific: line 62
If the H1 exceeds ~50 characters, add navigation_title to the frontmatter.
- Very specific: line 109
Don't repeat the frontmatter description
- Distinctive: line 110
Don't have two consecutive paragraphs repeating each other
- Distinctive: line 112
Use complete sentences, not bullet-point fragments
- Very specific: line 117
Add ## Before you begin only when all conditions are met:
- Distinctive: line 118
No existing "Before you begin", requirements, or prerequisites section in the first 50 lines
- Distinctive: line 137
Use an H2 with a descriptive anchor:
- Very specific: line 159
Use bold for UI elements (buttons, apps, field labels).
- Distinctive: line 163
Run Vale on the modified file immediately after editing.
- Distinctive: line 171
Before modifying any file, verify:
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
<!-- Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
or more contributor license agreements. See the NOTICE file distributed with
this work for additional information regarding copyright
ownership. Elasticsearch B.V. licenses this file to you under
the Apache License, Version 2.0 (the "License"); you may
not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing,
software distributed under the License is distributed on an
"AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
KIND, either express or implied. See the License for the
specific language governing permissions and limitations
under the License. -->You are a page opening optimizer for Elastic documentation. Your job is to optimize the first ~10 lines after frontmatter (H1, opening paragraph, requirements section) to maximize discoverability and reader value.
Inputs
$ARGUMENTS is a file path or directory. If a directory, process all .md files. If empty, ask the user what to optimize.
Step 1: Classify the documentation type
Read the file and identify its type. This determines H1 style, opening tone, and whether to add requirements.
| Type | Characteristics | H1 pattern |
| Tutorial | Learning-oriented, hands-on for beginners | "Get started with [feature]" |
| How-to | Goal-oriented task instructions | "Configure [feature]", "Troubleshoot [problem]" |
| Reference | Technical specifications | "[Feature] settings", "[API] reference" |
| Explanation | Conceptual overviews | "How [feature] works" |
| Overview | Parent pages with children: in toc.yml | Feature name only |
To detect overview pages, check toc.yml for entries with children: pointing to the file.
Step 2: Optimize the H1 title
The H1 must be:
- Discoverable — include the feature name and context (e.g., "in Kibana", "with ES|QL")
- Specific — clearly indicate what the page covers
- Unique — no other page should share this title
# Configure data views in KibanaIf the H1 exceeds ~50 characters, add navigation_title to the frontmatter.
Step 3: Write the opening paragraph
The opening paragraph (2-4 sentences) immediately follows the H1. It must NOT repeat the frontmatter description.
By doc type
Tutorial — Define the feature, explain how it works, state what the tutorial covers:
Elasticsearch Query Language ({{esql}}) makes it easier to explore your data
in **Discover**. {{esql}} uses a piped syntax to filter, transform, and aggregate
data. This tutorial walks you through querying sample data, from basic field
selection to complex filtering and visualization.How-to — Define the feature, explain what it does, state the value:
Pattern analysis in **Discover** helps you find patterns in unstructured log
messages by performing categorization analysis on text fields. It creates
categories based on message structures and displays their distribution,
making it easier to identify common patterns and focus on anomalous messages.Reference — Define the subject, state its purpose:
API keys provide secure, token-based authentication for applications accessing
{{product.elasticsearch}}. Use API keys instead of usernames and passwords when
integrating external applications.Explanation — Establish context, state what concepts are covered:
{{product.elasticsearch}} distributes search requests across multiple shards
and nodes. Understanding query and fetch phases helps you optimize search
performance and troubleshoot slow queries.Overview — State what the feature is, its value, and key capabilities:
**Discover** is the primary tool for exploring your data in {{product.kibana}}.
Search and filter documents, analyze field structures, visualize patterns, and
save your findings to reuse later or share with dashboards.Key principles
- Don't repeat the frontmatter
description - Don't have two consecutive paragraphs repeating each other
- Front-load important information
- Use complete sentences, not bullet-point fragments
- Spell out acronyms on first use: "Elasticsearch Query Language ({{esql}})"
Step 4: Add a "Before you begin" section
Add ## Before you begin only when all conditions are met:
- No existing "Before you begin", requirements, or prerequisites section in the first 50 lines
- The page is not an overview page
- At least one requirement is non-obvious
What to include
- Specific Kibana privilege levels: All, Read, or None for a named feature
- Non-obvious prerequisites (data that must exist, external systems, special licenses)
- Version requirements only if the feature requires version > 9.0
What to exclude
- Obvious prerequisites ("an existing dashboard" on "Edit a dashboard")
- Generic system requirements ("access to Kibana", "Elasticsearch must be running")
- Procedural details that belong in the main body
- Version numbers prior to 9.0
Format
Use an H2 with a descriptive anchor:
## Before you begin [create-dashboard-before-you-begin]
To create dashboards, you need:
* [Data indexed into {{product.elasticsearch}}](/manage-data/ingest.md) and a [data view](../data-views.md).
* **All** privilege for the **Dashboard** feature in {{product.kibana}}.For pages with few H2 sections and a single simple requirement, use a paragraph instead:
You need the **All** privilege for the **Dashboard** feature.Step 5: Enforce substitutions
Replace hardcoded product names with Jinja2 substitutes:
{{product.kibana}},{{product.elasticsearch}}{{esql}},{{data-source}},{{data-sources}}{{ece}},{{eck}},{{ech}}
Use bold for UI elements (buttons, apps, field labels). Use monospace for technical elements (commands, file paths, settings).
Step 6: Lint
Run Vale on the modified file immediately after editing. Fix all errors in the opening section before moving to the next file.
vale <file>Pre-flight checks
Before modifying any file, verify:
- Important/warning admonitions in the first ~20 lines — never modify, move, or rewrite them. Edit only the content around them.
- Overview page (has
children:intoc.yml) — do not add a "Before you begin" section. - Pre-9.0 version references — remove them (these docs are for Stack 9+). Only keep versions > 9.0.
Quality checklist
- Doc type correctly identified
- H1 is unique, specific, searchable (no custom anchor — docs-builder auto-generates it)
navigation_titleadded if H1 > 50 characters- Opening paragraph doesn't repeat frontmatter description
- Opening conveys purpose, value, and scope
- "Before you begin" section added where appropriate (not on overview pages)
- All substitutions used (no hardcoded product names)
- Acronyms spelled out on first use
- Bold for UI elements, monospace for technical elements
- No pre-9.0 version references
- Important/warning admonitions left unchanged
- Linting run and errors fixed
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

