Syncing Docs via CLI / GitHub

ReadMe's official command-line interface (CLI) and GitHub Action 🌊

If you're anything like us, your documentation process may be a part of a broader CI/CD process. For example, you may want to automatically update your Guides or API reference on ReadMe every time you ship new code. Enter rdme: ReadMe's official command-line interface (CLI) and GitHub Action!

With rdme, you can create workflows for a variety of use cases, including:

• Syncing OpenAPI/Swagger definitions (with support for bundling external references) 📦
• Pre-upload validation (including OpenAPI 3.1) ✅
• Syncing directories of Markdown files 📖

General Setup and Usage

To see detailed CLI setup instructions and all available commands, check out the rdme GitHub repository.

Markdown File Setup

📘

Guides, Changelog, Custom Pages... you name it!

The following guidance on Markdown file setup is nearly identical for Guides (i.e. the docs command), Changelog (i.e. the changelogs command), and Custom Pages (i.e. the custompages command). There are a couple of small differences:

• Guides are tied to project versions and therefore require a --version parameter. Changelog and Custom Pages are the same across project versions and therefore do not have a --version parameter.
• There are slight variations in the YAML front matter attributes for each respective section of your documentation. For example, Changelog has a type attribute which you can set to added. See Specifying Other Attributes for more information.
• In addition to Markdown, Custom Pages also supports HTML files. If you pass an HTML file into the custompages commands, the page will have the htmlmode flag set to true and it will conversely be set to false for Markdown files. You can override this in the YAML front matter.

In order to sync Markdown files to your Guides, your Changelog, or your Custom Pages, you'll need to add certain attributes to the top of each page via a YAML front matter block. See below for an example (using the page you're currently reading!):

---
title: Syncing Docs via CLI / GitHub
excerpt: Update your docs automatically with `rdme`, ReadMe's official CLI and GitHub Action!
category: 62056dee230e07007218be06
---

If you're anything like us...

Required Attributes

See below for a table detailing the required YAML front matter attributes:

To determine what your category value should be, you can use the Get all categories endpoint and grab the id value from the response.

📘

Any Markdown/HTML files that lack YAML front matter attributes will be skipped.

Specifying Page Slugs

By default, we automatically derive the page's slug via the file name (e.g. the file name rdme.md would become /docs/rdme in your ReadMe project). Note that our API uses slugify to automatically handle certain characters (e.g. spaces), which may lead to unexpected syncing behavior if your file names don't match your page slugs. If you prefer to keep your page slugs different from your file names, you can manually set the slug value in the YAML front matter:

---
title: Syncing Docs via CLI / GitHub
excerpt: Update your docs automatically with `rdme`, ReadMe's official CLI and GitHub Action!
category: 62056dee230e07007218be06
slug: an-alternative-page-slug-example
---

Specifying Other Attributes

You can also specify several other page attributes in your YAML front matter, such as hidden (a boolean which denotes whether your page is published or unpublished). Any attributes you omit will remain unchanged on rdme runs. To view the full list of attributes, check out the POST endpoints for respective section of your documentation that you're syncing:

• Create doc
• Create changelog
• Create custom page

Dry Run Mode

If you're setting up new pages or if you're generally unsure if you've set up your page attributes correctly, each command has a dry run mode. This will allow you preview the changes without actually creating/updating any docs in ReadMe, which can be extremely useful for initial setup (oh, and we have comprehensive debugging options available as well!). To enable dry run mode, use the --dryRun flag:

rdme docs [path] --version={project-version} --dryRun
rdme changelogs [path] --dryRun
rdme custompages [path] --dryRun

The command output will indicate whether each page is being created or updated alongside all processed page attributes.

GitHub Actions Usage

With GitHub Actions, you can automatically execute workflows when certain events take place in your GitHub repository (e.g. code is pushed to the default branch, a new pull request is opened, etc.).

While there are dozens of event options available, you'll typically want to sync your OpenAPI definition and Markdown docs to ReadMe when one of the following events takes place:

• Code is pushed to the default branch 🌴
• A build is deployed 🚀
• A release is created 🆕

📘

Keeping rdme up-to-date

Note that @v8 (used in the examples on this page) is the latest version of rdme. We recommend configuring Dependabot to keep your actions up-to-date.

Quick Start

The fastest way to create a GitHub Actions workflow file is by doing the following:

1. Download rdme to your local machine (see our setup instructions)
2. Open your command line at the root of your GitHub repository
3. Run the command you wish to automate on your local machine with the --github flag tacked on at the end. For example:

rdme openapi --github

This will run through the openapi command, ask you a few quick questions, and then automatically create a fully functional GitHub Actions workflow file for you. 🪄

👍

Did you know?

Not to get too meta on you, but... the page that you're currently reading is actually being synced from the rdme GitHub repository via the rdme GitHub Action 🤯

Here are the relevant files on GitHub:

• The Markdown source file for the page you're reading 📜
• The GitHub Actions workflow file that syncs the Markdown to docs.readme.com 🔄
• And finally... the workflow run results ✅

Securely Using Your API Key

🚧

Secretly store your ReadMe API Key!

GitHub Actions has secrets to securely store sensitive information so it isn't publicly visible. We strongly recommend using these for storing your ReadMe API Key and any other secret keys—whether your repository is public or private. You can read more about setting these up in their documentation.

To use sensitive information (like your ReadMe API key) in your rdme GitHub Action, first create a new repository secret. Let's say you create a new secret key called README_API_KEY. The usage in the rdme step will look something like this:

- uses: readmeio/[email protected]
  with:
    rdme: openapi [url-or-local-path-to-file] --key=${{ secrets.README_API_KEY }} --id=API_DEFINITION_ID

Manually Configuring a GitHub Actions Workflow

We recommend using the quick start to get started with GitHub Actions. But if you are familiar with the workflow syntax and wish to manually configure a workflow, follow these steps:

1. Create a new GitHub Actions workflow file in the .github/workflows directory of your repository (or use an existing workflow file)

2. Configure the on property, which determines what triggers the execution of the workflow.

3. Add the following two steps to your workflow:

# Required in order for the GitHub Action to access your repo's contents
- uses: actions/[email protected]

# Runs the `rdme` command on your repo's contents
- uses: readmeio/[email protected]
  with:
    rdme: [your command here]

The command syntax in GitHub Actions is functionally equivalent to the CLI. For example, take the following CLI command:

rdme validate [url-or-local-path-to-file]

To execute this command via GitHub Actions, the step would look like this:

- uses: readmeio/[email protected]
  with:
    rdme: validate [url-or-local-path-to-file]

The following section has links to full GitHub Actions workflow file examples.

GitHub Actions Examples

Want to start syncing? We have several example workflow files available:

• Syncing an OpenAPI definition
• Syncing a directory of Markdown files

Usage in Other CI Environments

Since rdme is a command-line tool at its core, you can use rdme to sync your documentation from virtually any CI/CD environment that runs shell commands—Bitbucket Pipelines, GitLab CI/CD, you name it! You can do this by invoking rdme with npx [email protected] in a Node.js environment. See below for several examples.

# Official framework image. Look for the different tagged releases at:
# https://hub.docker.com/r/library/node/tags/
image: node:14
pipelines:
  default:
    - step:
        script:
          - npx [email protected] openapi [url-or-local-path-to-file] --key=$README_API_KEY --id=API_DEFINITION_ID

version: 2.1
jobs:
  sync-via-rdme:
    docker:
      # Official framework image. Look for the different tagged releases at:
      # https://hub.docker.com/r/library/node/tags/
      - image: node:14
    steps:
      - run:
          command: npx [email protected] openapi [url-or-local-path-to-file] --key=$README_API_KEY --id=API_DEFINITION_ID

# Official framework image. Look for the different tagged releases at:
# https://hub.docker.com/r/library/node/tags/
image: node:14

sync-via-rdme:
  script:
    - npx [email protected] openapi [url-or-local-path-to-file] --key=$README_API_KEY --id=API_DEFINITION_ID

# https://docs.travis-ci.com/user/languages/javascript-with-nodejs/#specifying-nodejs-versions
language: node_js
node_js:
  - 14

script: npx [email protected] openapi [url-or-local-path-to-file] --key=$README_API_KEY --id=API_DEFINITION_ID

If you notice any issues with any of these examples, please open up an issue on the rdme repository on GitHub.

🚧

Secretly store your ReadMe API Key!

Nearly every CI service has a way to securely add secrets so that they're not exposed in your scripts and build logs. We strongly recommend using such a feature for storing your ReadMe API key. The examples above use $README_API_KEY, which is how you typically load such variables in your scripts. We've included some links below on how to configure these for the respective examples:

• Bitbucket Pipelines
• CircleCI
• GitLab CI
• Travis CI

Troubleshooting

If you're running into unexpected behavior with rdme and need to troubleshoot the issue, you have several debug logging options available. We may ask for these logs (as well as a copy of your OpenAPI definition) when you contact our support team.

If you're working with the docs command specifically, we recommend using dry run mode first so your docs don't get overwritten. If you're still seeing unexpected results (or are running into issues with any other command), check out the debugging options described below.

Troubleshooting CLI

If you're troubleshooting issues with the CLI (or in some non-GitHub Actions environment), you can use the DEBUG environmental variable to print helpful debugging info to the console:

DEBUG=rdme* rdme openapi [url-or-local-path-to-file]

Note that this should only be used for development/debugging purposes and should not be enabled in production environments.

Troubleshooting GitHub Actions

If you're troubleshooting issues in a GitHub Actions environment, you can enable step debug logs in your GitHub Actions workflow by setting the repository secret ACTIONS_STEP_DEBUG to true. For more information on accessing, downloading, and deleting logs, check out GitHub's documentation.

🚧

Debug Logs May Contain Sensitive Information

Enabling step debug logs will produce comprehensive logging for all of your GitHub Actions workflow steps. While GitHub automatically masks any sensitive information you load in with secrets, there might be other sensitive information that's exposed. Anybody with read access to the repository will be able to see these logs.

We strongly recommend that you only enable step debug logs in private repositories. If working in a public repository, we suggest recreating your GitHub workflow setup (e.g. with your GitHub workflow files, OpenAPI definitions, and anything else you need for syncing to ReadMe) in a separate private repository for testing purposes before enabling this setting.

If you do enable step debug logs in your repository and your logs produce sensitive information, here are GitHub's docs on deleting logs.