[No QA] [Hold for feature] [No QA] Update help docs for Release 3: Custom Agents

[MelvinBot]

@stephanieelliott note: Design request for images here, no need to block article publishing on these!

Summary

Updates the help site for Release 3 of the Custom Agents project — the first customer-facing release, which promotes agents inside workspace Approvals and Rules.

What's new

• New article: docs/articles/new-expensify/agents/Custom-Agents.md — explains what Custom Agents are, how to create one (from Settings > Agents or inline from a workflow), how to write agent instructions, how to chat with and Copilot into an agent, how to edit/delete an agent, and how to use an agent as an approver.
• New help hub: Adds an AI Agents hub to docs/_data/_routes.yml under the New Expensify section, pointing at /new-expensify/hubs/ai-agents/.
• Pointer added to docs/articles/new-expensify/workspaces/Add-Approvals.md — short Set a Custom Agent as Approver subsection under Configure Approval Workflows, with a link to the new article.
• Pointer added to docs/articles/new-expensify/workspaces/Workspace-Rules.md — one line under the overview noting that an agent can be used as an approver when approval logic doesn't fit the standard rules engine.

Context

Scoping issue: Expensify/Expensify#642268

This PR follows the Release 3 scope outlined in the resource updates issue: one new landing article and two minor pointer additions. The agent feature is additive — it does not change how existing approvals or rules work — so no broader revisions are needed.

Billing details and personal agent usage are intentionally deferred to Releases 4 and 5.

Test plan

• Render the help site locally and confirm the new Agents hub appears under New Expensify.
• Confirm the new Custom-Agents.md article renders at /new-expensify/hubs/agents/Custom-Agents.
• Confirm the new pointers in Add-Approvals.md and Workspace-Rules.md render with working links to the new article.
• Verify the article copy matches the actual UI labels in the product before merging (e.g., Settings > Agents, agent+ button, Edit / Chat / Copilot / Delete buttons, AI Prompt page name).

i guess i need to add the checklist

PR Author Checklist

• I linked the correct issue in the ### Fixed Issues section above
• I wrote clear testing steps that cover the changes made in this PR
  • I added steps for local testing in the Tests section
  • I added steps for the expected offline behavior in the Offline steps section
  • I added steps for Staging and/or Production testing in the QA steps section
  • I added steps to cover failure scenarios (i.e. verify an input displays the correct error message if the entered data is not correct)
  • I turned off my network connection and tested it while offline to ensure it matches the expected behavior (i.e. verify the default avatar icon is displayed if app is offline)
  • I tested this PR with a High Traffic account against the staging or production API to ensure there are no regressions (e.g. long loading states that impact usability).
• I included screenshots or videos for tests on all platforms
• I ran the tests on all platforms & verified they passed on:
  • Android: Native
  • Android: mWeb Chrome
  • iOS: Native
  • iOS: mWeb Safari
  • MacOS: Chrome / Safari
• I verified there are no console errors (if there's a console error not related to the PR, report it or open an issue for it to be fixed)
• I followed proper code patterns (see Reviewing the code)
  • I verified that any callback methods that were added or modified are named for what the method does and never what callback they handle (i.e. toggleReport and not onIconClick)
  • I verified that comments were added to code that is not self explanatory
  • I verified that any new or modified comments were clear, correct English, and explained "why" the code was doing something instead of only explaining "what" the code was doing.
  • I verified any copy / text shown in the product is localized by adding it to src/languages/* files and using the translation method
  • I verified all numbers, amounts, dates and phone numbers shown in the product are using the localization methods
  • I verified any copy / text that was added to the app is grammatically correct in English. It adheres to proper capitalization guidelines (note: only the first word of header/labels should be capitalized), and is either coming verbatim from figma or has been approved by marketing (in order to get marketing approval, ask the Bug Zero team member to add the Waiting for copy label to the issue)
  • I verified proper file naming conventions were followed for any new files or renamed files. All non-platform specific files are named after what they export and are not named "index.js". All platform-specific files are named for the platform the code supports as outlined in the README.
  • I verified the JSDocs style guidelines (in STYLE.md) were followed
• If a new code pattern is added I verified it was agreed to be used by multiple Expensify engineers
• I followed the guidelines as stated in the Review Guidelines
• I tested other components that can be impacted by my changes (i.e. if the PR modifies a shared library or component like Avatar, I verified the components using Avatar are working as expected)
• I verified all code is DRY (the PR doesn't include any logic written more than once, with the exception of tests)
• I verified any variables that can be defined as constants (ie. in CONST.ts or at the top of the file that uses the constant) are defined as such
• I verified that if a function's arguments changed that all usages have also been updated correctly
• If any new file was added I verified that:
  • The file has a description of what it does and/or why is needed at the top of the file if the code is not self explanatory
• If a new CSS style is added I verified that:
  • A similar style doesn't already exist
  • The style can't be created with an existing StyleUtils function (i.e. StyleUtils.getBackgroundAndBorderStyle(theme.componentBG))
• If new assets were added or existing ones were modified, I verified that:
  • The assets are optimized and compressed (for SVG files, run npm run compress-svg)
  • The assets load correctly across all supported platforms.
• If the PR modifies code that runs when editing or sending messages, I tested and verified there is no unexpected behavior for all supported markdown - URLs, single line code, code blocks, quotes, headings, bold, strikethrough, and italic.
• If the PR modifies a generic component, I tested and verified that those changes do not break usages of that component in the rest of the App (i.e. if a shared library or component like Avatar is modified, I verified that Avatar is working as expected in all cases)
• If the PR modifies a component related to any of the existing Storybook stories, I tested and verified all stories for that component are still working as expected.
• If the PR modifies a component or page that can be accessed by a direct deeplink, I verified that the code functions as expected when the deeplink is used - from a logged in and logged out account.
• If the PR modifies the UI (e.g. new buttons, new UI components, changing the padding/spacing/sizing, moving components, etc) or modifies the form input styles:
  • I verified that all the inputs inside a form are aligned with each other.
  • I added Design label and/or tagged @Expensify/design so the design team can review the changes.
• If a new page is added, I verified it's using the ScrollView component to make it scrollable when more elements are added to the page.
• I added unit tests for any new feature or bug fix in this PR to help automatically prevent regressions in this user flow.
• If the main branch was merged into this PR after a review, I tested again and verified the outcome was still expected according to the Test steps.

Screenshots/Videos

[melvin-bot]

Concierge reviewer checklist:

• I have verified the accuracy of the article
  • The article is within a hub that makes sense, and the navigation is correct
  • All processes, screenshots, and examples align with current product behavior.
  • All links within the doc have been verified for correct destination and functionality.
• I have verified the readability of the article
  • The article's language is clear, concise, and free of jargon.
  • The grammar, spelling, and punctuation are correct.
  • The article contains at least one image, or that an image is not necessary
• I have verified the formatting of the article
  • The article has consistent formatting (headings, bullet points, etc.) with other HelpDot articles and that aligns with the HelpDot formatting standards.
  • The article has proper text spacing and breaks for readability.
• I have verified the article has the appropriate tone and style
  • The article's tone is professional, friendly, and suitable for the target audience.
  • The article's tone, terminology, and voice are consistent throughout.
• I have verified the overall quality of the article
  • The article is not missing any information, nor does it contain redundant information.
  • The article fully addresses user needs.
• I have verified that all requested improvements have been addressed

For more detailed instructions on completing this checklist, see How do I review a HelpDot PR as a Concierge Team member?

cc @strepanier03

[melvin-bot]

@Beamanator @strepanier03 One of you needs to copy/paste the Reviewer Checklist from here into a new comment on this PR and complete it. If you have the K2 extension, you can simply click: [this button]

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: f0890d28d4

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[MelvinBot]

The failing checks aren't related to the PR's code:

• checklist (×2) — PR Author Checklist and PR Reviewer Checklist are expected to fail until the human co-author ticks the items off. These are on the bot's ignore list.
• review (3m) — the PR Reviews with Claude Code workflow run (triggered by pull_request_target) was cancelled mid-run, not failed by content. A second run of the same workflow on the same commit (f0890d2, triggered by pull_request_review) completed successfully. Nothing to fix here.

No push needed.

[github-actions]

HelpDot Documentation Review

Overall Assessment

This PR introduces a new "AI Agents" hub and a comprehensive article on creating and using Custom Agents, along with cross-link additions to two existing workspace articles (Add-Approvals.md and Workspace-Rules.md). The new article is well-organized, covers the workflow thoroughly, and follows most governance conventions. However, there are two broken cross-links and a few heading and navigation phrasing issues that should be resolved before merge.

Scores Summary

• Readability: 8/10 - Clear step formatting, logical section flow, and a useful example prompt block. The repeated navigation preamble across five sections is slightly verbose but acceptable for standalone scanability.
• AI Readiness: 9/10 - Strong YAML metadata including internalScope, descriptive keywords, and task-based "How to..." headings throughout. The H1/H2-only structure is correctly followed, and the FAQ section is well-formed.
• Style Compliance: 6/10 - Two broken cross-link URLs, one non-task-based heading in the Add-Approvals.md addition, and a minor navigation phrasing deviation from the governance standard.

Key Findings

Critical -- must fix:

• Broken cross-links in two files. Both Add-Approvals.md and Workspace-Rules.md link to /articles/new-expensify/ai-agents/docs/articles/Create-and-Use-Custom-Agents. The path contains a duplicated docs/articles segment. The correct relative link should be /articles/new-expensify/ai-agents/Create-and-Use-Custom-Agents.

Should fix:

• Non-task-based heading. The new section in Add-Approvals.md uses ## Set a Custom Agent as Approver. Per HELP_AUTHORING_GUIDELINES.md Section 2, all ## headings must start with an action verb or question word in a task-based phrase. Suggested fix: ## How to set a Custom Agent as an approver.
• Navigation phrasing. The unified navigation instruction throughout the new article reads: "In the navigation tabs (on the left on web, and at the bottom on mobile)". The governance standard in HELPSITE_NAMING_CONVENTIONS.md specifies: "In the navigation tabs (on the left on web, on the bottom on mobile)" -- without "and at the". This appears six times in the new article.

Positive aspects:

• The YAML frontmatter is thorough, with a well-scoped internalScope field that clearly defines audience and exclusions.
• The article follows the one-workflow-per-article principle while covering create, edit, delete, chat, copilot, and approver assignment as logical sub-tasks of a single feature.
• FAQ questions are task-based and directly address likely user concerns.
• Cross-links use descriptive anchor text (e.g., "Learn how to..." and "Learn more about...") per governance standards.
• Screenshot placeholders follow the required HTML comment format.
• Bold formatting on UI elements (button names, tab names) is consistently applied.

Recommendations

1. Fix the broken links in Add-Approvals.md and Workspace-Rules.md -- remove the extra docs/articles/ from the path.
2. Rename the heading in Add-Approvals.md from ## Set a Custom Agent as Approver to a task-based format such as ## How to set a Custom Agent as an approver.
3. Update the navigation phrasing in the new article to match the governance standard exactly: replace "and at the bottom on mobile" with "on the bottom on mobile" in all six occurrences.

Files Reviewed

• docs/_data/_routes.yml -- New hub entry added; no issues.
• docs/articles/new-expensify/ai-agents/Create-and-Use-Custom-Agents.md -- New article; navigation phrasing deviation (minor).
• docs/articles/new-expensify/workspaces/Add-Approvals.md -- Broken cross-link URL and non-task-based heading.
• docs/articles/new-expensify/workspaces/Workspace-Rules.md -- Broken cross-link URL.
• docs/new-expensify/hubs/ai-agents/index.html -- New hub page; no issues.

[MelvinBot]

The failing checks aren't actionable — no code fix is needed:

• PR Author Checklist and PR Reviewer Checklist are expected to fail until the human co-author / a reviewer fills out the respective sections in the PR body.
• The failing review entry is a stale, canceled run of PR Reviews with Claude Code (26672734346). Two later runs of the same workflow on the same commit (26672795713 and 26673129051) both passed. The display just hasn't dropped the canceled run.

All other checks are passing.

[OSBotify]

A preview of your ExpensifyHelp changes have been deployed to https://883d0dcb.helpdot.pages.dev ⚡️

Updated articles:

• articles/new-expensify/ai-agents/Create-and-Use-Custom-Agents
• articles/new-expensify/workspaces/Add-Approvals
• articles/new-expensify/workspaces/Workspace-Rules

[stephanieelliott]

This is ready if you want to give it a review @yuwenmemon! Couple notes:

• I have a separate issue open to get screenshots from Design - I'll add them once we have them.
• The original suggested hub name (new-expensify/agents) felt too short and vague against the other hub names. I named it new-expensify/ai-agents) instead but LMK if we need to use the first one or there is a better name to use. FWIW the hub name is what will appear on this page
• I added an example prompt in the article here, I have no idea whether that is the optimal prompt format, so please check that if you don't mind!