agentleFS
igmarinigmarin/rails-agent-skills/skills/implement-authorization/SKILL.md

implement-authorization

Skill26 starsChanged 4 months agoMITView on GitHub

Use when adding or reviewing Rails authorization with Pundit, CanCanCan, or policy objects. Trigger words: authorization, Pundit, CanCanCan, policy, roles, permissions.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    When implementing or reviewing authorization, the output answer.md must include:

    line 101

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Implement Authorization

Quick Reference

GemPatternBest For
PunditExplicit policy classesComplex per-resource rules
CanCanCanCentralized Ability classSimple role-based permissions

Core Process

Implementation Workflow

  1. Inspect existing authorization — reuse the installed policy framework; add pundit or cancancan only if the task requires a new framework and project conventions permit it
  2. Generate base only when needed — run the gem's installer (rails g pundit:install or rails g cancan:ability) only when no supported framework is present and the project permits the generator; otherwise proceed with the detected framework without invoking generators
  3. Define policies/abilities — create policy classes (Pundit) or populate the Ability class (CanCanCan); always use policy objects, never inline authorization logic in controllers
  4. Authorize in controllers — call authorize @record (Pundit) or authorize! :action, @record (CanCanCan) in each action
  5. Verify authorization — attempt an unauthorized action in the browser or console and confirm it raises Pundit::NotAuthorizedError or CanCan::AccessDenied as expected; use persisted records (e.g., User.create!) not unsaved ones
  6. Scope queries — use policy_scope(Model) or accessible_by(current_ability) for index actions
  7. Test all roles — write policy specs and request specs covering admin, owner, and guest; check specific permissions, never presence checks alone

Patterns

Pundit

class PostPolicy < ApplicationPolicy
  def update?
    user.admin? || record.user_id == user.id
  end
end

CanCanCan

class Ability
  include CanCan::Ability

  def initialize(user)
    can :update, Post, user_id: user.id
    can :manage, :all if user.admin?
  end
end

Troubleshooting

ErrorLikely CauseFix
Pundit::NotDefinedErrorNo policy class found for the recordCreate app/policies/model_policy.rb inheriting from ApplicationPolicy
Pundit::AuthorizationNotPerformedErrorauthorize not called in a controller actionAdd authorize @record in the action, or after_action :verify_authorized to catch misses
CanCan::AccessDenied unexpectedly raisedAbility rules not matching the current user/roleInspect current_ability.can?(:action, @record) in the console to debug rule evaluation

Testing

Cover every role (admin, owner, guest) in both policy specs and request specs.

Minimal Pundit policy spec

RSpec.describe PostPolicy do
  subject { described_class.new(user, post) }

  let(:post) { create(:post, user: owner) }
  let(:owner) { create(:user) }

  context 'as admin' do
    let(:user) { create(:user, :admin) }
    it { is_expected.to permit_action(:update) }
  end

  context 'as owner' do
    let(:user) { owner }
    it { is_expected.to permit_action(:update) }
  end

  context 'as guest' do
    let(:user) { create(:user) }
    it { is_expected.not_to permit_action(:update) }
  end
end

Output Style

When implementing or reviewing authorization, the output answer.md must include:

  1. Manual Denied-Action Verification — a dedicated section with actual Rails console output, or an explicit unrun status showing the authorization exception raised when an unauthorized action is attempted. Always use persisted records (User.create!, Post.create!), never unsaved ones.
  2. HTTP and Policy Verification — concrete curl requests or controller test commands with expected HTTP response codes (e.g. 403 Forbidden or 302 Found) when access is denied.
  3. Language — English unless explicitly requested otherwise.

See references/output-style.md for full formatting examples including Pundit and CanCanCan console output templates.

Integration

SkillWhen to chain
write-testsWhen implementing authorization tests.

Extended Resources (Progressive Disclosure)

Load these files only when their specific content is needed:

  • EXAMPLES.md — Use when you need complete Pundit or CanCanCan implementation examples beyond the inline samples
  • references/workflow.md — Use when you need the step-by-step authorization implementation workflow diagram
  • references/output-style.md — Use when you need full formatting templates for console verification output

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.