Auditing Workflow Conventions logo

Auditing Workflow Conventions

Organization
bitwarden
auditing-workflow-conventions

Reference for Bitwarden GitHub Actions naming conventions that the workflow linter (bwwl) does not enforce. Covers three standards — job IDs (kebab-case), step names (Sentence case imperative), and workflow file names (kebab-case.yml, `_` prefix for reusable) — plus an advisory canonical step-name glossary and reference-sweep procedures for job ID and filename renames. Use when auditing or authoring workflows and questions like "what casing should job IDs use", "should this reusable workflow be build.yml or _build.yml", or "review these workflows for naming consistency" come up. Read alongside bitwarden-workflow-linter-rules, which is the source of truth for linted rules; this skill covers only the gaps.

Overview

Publisherbitwarden
Repositoryai-plugins
Skill nameauditing-workflow-conventions
Stars
149
Forks
19
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by bitwarden on GitHub. Read the source before you install it.

Installation

Install the Auditing Workflow Conventions AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/bitwarden/ai-plugins.git /tmp/ai-plugins
mkdir -p .claude/skills
cp -r /tmp/ai-plugins/plugins/bitwarden-devops-engineer/skills/auditing-workflow-conventions .claude/skills/auditing-workflow-conventions
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Auditing Workflow Conventions in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Auditing Workflow Conventions on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Auditing Workflow Conventions is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Ownership

This skill covers only what bwwl cannot check. For anything the linter enforces, invoke Skill(bitwarden-devops-engineer:bitwarden-workflow-linter-rules) — that skill is the source of truth for all bwwl rules, including their triggers and fix procedures. Do not report a finding here that duplicates a linter rule.

Naming categoryOwner
Job IDsThis skill
Step name casingThis skill
Workflow file namesThis skill
Workflow and job display name:bitwarden-workflow-linter-rulesname_capitalized, name_exists
Outputsbitwarden-workflow-linter-rulesunderscore_outputs
Job-level env varsbitwarden-workflow-linter-rulesjob_environment_prefix
Inputs, bash variables, artifact namesNo standard. Do not invent one — flag the gap instead.

name_capitalized checks only that the first character is capitalized. Get Package Version passes the linter and still deviates from the Sentence case standard below. The two are complementary.

Standards

These three IDs are defined by this skill, not by bwwl. Label them as convention findings, never as linter findings.

job-id-kebab-case

  • Applies to: every key under jobs:.
  • Standard: kebab-case.
  • Correct: build, test, lint, publish, upload, build-artifacts, deploy-service, check-permissions, calculate-version.
  • Deviates: bump_version, cut_branch, deployService, DeployService, BUILD_ARTIFACTS.
  • Flag when: a job ID uses snake_case, camelCase, PascalCase, or UPPER_SNAKE. Single-word IDs are compliant and are never a finding.

A job ID is an identifier, not a label. Before proposing a rename, resolve every reference:

  1. needs: in the same file, in both scalar and list form.
  2. ${{ needs.<job-id>.* }} expressions anywhere in the file.
  3. jobs.<job-id>.outputs at the workflow level of a reusable workflow.

Check-run names derive from a job's name:, not its ID, so a job-ID rename does not affect required status checks — unless the job has no name:, in which case the ID becomes the check name and a rename can break a ruleset that requires it. name_exists already flags a job with no name:; if you encounter one, fix that first and rename second.

step-name-sentence-case

  • Applies to: every name: under steps:.
  • Standard: Sentence case with a leading imperative verb — capitalize the first word plus proper nouns and acronyms only. Proper nouns and acronyms keep their own casing: Azure, Docker, Node, .NET, SDK, RC, MSSQL.
  • Correct: Check out repo, Set up .NET, Print environment, Log in to Azure, Generate Docker image tag, Install Node dependencies, Run tests, Push changes, Upload SDK artifacts.
  • Deviates: Delete Release Branch, Get Package Version, Build & Package Binaries, Upload SDK Artifacts, Dependency install.
  • Flag when: a step name uses Title Case, ALL CAPS, or drops the leading imperative verb.

Step names are display-only — they are not addressable from expressions, since steps.<id> resolves the step's id:. Changing one is safe and needs no reference sweep. This is the only standard here an agent may apply directly during an edit.

workflow-file-naming

  • Applies to: every file in .github/workflows/.
  • Standard: kebab-case.yml, with a _ prefix if and only if the workflow is exclusively reusable. The extension is always .yml, never .yaml. The prefix and the casing are independent — _deploy_service.yml is correctly prefixed and incorrectly cased; the compliant form is _deploy-service.yml.
  • Correct: build-app.yml, scan-dependencies.yml, _version.yml, _deploy-service.yml.
  • Deviates: build_only.yml, API_tests.yml, Integration_Tests.yml, deploy-service.yaml, _deploy_service.yml.
  • Flag when: the name is not kebab-case, the extension is .yaml, an exclusively reusable workflow lacks the _ prefix, or a _-prefixed file is not exclusively reusable.

Exclusively reusable means the on: block declares nothing outside workflow_call and workflow_dispatch. workflow_dispatch is a manual testing and operations escape hatch, not a standalone entry point, so it does not disqualify the prefix.

Any other trigger — push, pull_request, pull_request_target, schedule, release, workflow_run, repository_dispatch — makes the workflow dual-purpose: it both runs on its own and is callable by others. Dual-purpose workflows take no prefix. This is a deliberate and common design, not a deviation. Presence of workflow_call alone never justifies a prefix finding; check the full trigger set first.

on: blockPrefix
workflow_callRequired
workflow_call + workflow_dispatchRequired
workflow_call + any auto-triggerNone — dual-purpose
No workflow_callNone

Renaming a workflow file is destructive. The filename is the workflow's public identifier — it addresses the workflow from outside the repo, and every external reference breaks silently on rename. Never rename one as part of an audit or an unrelated edit. Report it and let the repo owner schedule it.

If a rename is explicitly requested, establish the known reference set first, then use git mv:

  1. In-repo callers — grep the repo for the filename. Catches uses: ./.github/workflows/<file> and any local script that names it.
  2. Org-wide referencesgh search code "<file>" --owner bitwarden. Catches uses: bitwarden/<repo>/.github/workflows/<file>@<ref> in other repos plus any script or tooling that names the file. A local grep will not find these.

This establishes a floor, not a ceiling. Nothing outside the org's indexed code is discoverable: automation in other systems, runbook and documentation links, and anything invoking the workflow by filename through the dispatch API. Report what the two searches found and state plainly that the set may be incomplete — the residual unknown is exactly why the rename belongs to the repo owner and not to an audit.

Run history is keyed to the file path and always detaches on rename. Prior runs remain but no longer group under the renamed workflow. This is unavoidable — surface it, do not try to preserve it.

Advisory: Canonical Step Names

Preferred wording, not a standard. Consistent phrasing makes steps greppable across repos, but a step name outside this table is never a finding — most step names are legitimately unique to their workflow.

ActionCanonical formAlso seen
Check out the repositoryCheck out repoCheckout repo, Checkout code, Checkout Branch, Check out repository, Checkout
Log in to AzureLog in to AzureLogin to Azure, Azure Login
Log out from AzureLog out from Azure
Install a toolchainSet up {tool} — two words, matching actions/setup-*Setup {tool}
Retrieve secrets from AKVRetrieve secretsGet secrets, Setup secrets
Print the environmentPrint environmentPrint Environment

When a step name both deviates from Sentence case and appears in this table (Azure Login), the casing is the finding and the canonical form is the suggested fix.

Applying These Standards

  • A deviation is a flag, not a verdict. These standards describe the target state; they do not authorize a rename. Real workflows carry deliberate exceptions — a job ID matched by external tooling, a filename referenced by a system outside this repo, a step name that reads better than the canonical phrase. Surface the deviation with what a compliant form would be, and let the owner decide. Do not treat silence as consent.
  • Only step names are safe to change in place. Job IDs and filenames both require a reference sweep first. Never fold either rename into an unrelated change.
  • This skill does not edit. Its allowed-tools is read-only — enough to inspect workflows and run the org-wide reference sweep, nothing more.
  • Renames are all-or-nothing. A partially applied rename is worse than the original deviation — it produces a broken workflow instead of an inconsistent one. If the full reference set cannot be resolved, do not start.
  • Do not double-report. If an item is already covered by name_capitalized, underscore_outputs, or job_environment_prefix, the finding belongs to the linter. See the ownership table.
  • Flag gaps honestly. Inputs, bash variable casing, and artifact names have no enforced standard. Say so; do not assert a convention this skill does not define.

Frequently asked questions

What does the Auditing Workflow Conventions AI skill do?

Reference for Bitwarden GitHub Actions naming conventions that the workflow linter (bwwl) does not enforce. Covers three standards — job IDs (kebab-case), step names (Sentence case imperative), and workflow file names (kebab-case.yml, `_` prefix for reusable) — plus an advisory canonical step-name glossary and reference-sweep procedures for job ID and filename renames. Use when auditing or authoring workflows and questions like "what casing should job IDs use", "should this reusable workflow be build.yml or _build.yml", or "review these workflows for naming consistency" come up. Read alongs...

Why use Auditing Workflow Conventions on TypingMind?

Because you install it once and use it with any model. Auditing Workflow Conventions is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Auditing Workflow Conventions in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/bitwarden/ai-plugins/tree/main/plugins/bitwarden-devops-engineer/skills/auditing-workflow-conventions. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Auditing Workflow Conventions?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Auditing Workflow Conventions?

As many as you like. As long as a model supports skills, you can use Auditing Workflow Conventions with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Auditing Workflow Conventions AI skill free?

It is published on GitHub by bitwarden. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇