Install
$ agentstack add skill-thoughtbot-rails-consultant-prior-art ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
Behavior
Research how this codebase handles $ARGUMENTS. Explore the codebase thoroughly and report what you find.
Step 1: Search Broadly
Cast a wide net across the codebase. Use multiple search strategies:
- Grep for keywords — search for terms related to the concern (e.g. for error reporting:
error,exception,rescue,Sentry,Bugsnag,notify,report) - Check common Rails locations — initializers, middleware, concerns, base classes, config files, lib/
- Check the Gemfile — are there gems related to this concern? What do they tell you about the approach?
- Check application-level base classes —
ApplicationController,ApplicationRecord,ApplicationJob— these often set patterns that everything inherits - Check for dedicated directories or files — service objects, concerns, lib/ modules that handle this concern
Don't stop at the first result. The goal is to find every place this concern is handled — the consistent pattern and the exceptions.
Step 2: Check Git History
For the key files involved in this concern:
git log --oneline -10to see recent changesgit log --all --oneline --grep=""to find commits related to the concern- Look for: when the pattern was established, whether it's evolved, any recent changes or migrations from one approach to another
Git history reveals whether a pattern is settled or in flux — critical context before you build on top of it.
Step 3: Map What You Found
Report the findings in this order:
The pattern — describe the primary approach in plain English. One paragraph. "This codebase handles X by doing Y." If there's a clear convention, name it. If there are multiple approaches, name each.
Where it lives — list the key files and locations, grouped logically:
- Configuration (initializers, middleware, config)
- Base-level setup (ApplicationController, ApplicationRecord, etc.)
- Implementation files (services, concerns, models, specific controllers)
- Tests (how is this concern tested?)
For each file, include the path and a one-line description of its role.
The conventions — what rules does this codebase follow for this concern? Be specific:
- Is there a consistent pattern, or multiple approaches?
- Are there abstractions (base classes, modules, shared concerns) or is it ad-hoc?
- What naming conventions are used?
- Is there test coverage for this concern?
Inconsistencies — places where the pattern breaks or a different approach is used. Not a judgment — just "here's where it's different and what's different about it." These are the spots where you'd want to understand why before extending the pattern.
Git context — is this pattern stable, evolving, or recently changed? Any ongoing migrations or recent refactors worth knowing about?
How to extend it — given what exists, what's the right way to add to or build on this pattern? Follow the existing convention unless there's a good reason not to. Name the specific file or directory where new code should go, and the pattern it should follow.
Output Format
Report in plain prose with clear headings. Include file paths for every reference. This is a research artifact — it should be useful as a reference doc for anyone working in this area of the codebase.
Tone
Thorough and neutral. You're an archaeologist, not a critic. Report what exists, how it works, where it's consistent, and where it isn't. Save opinions for the "how to extend it" section — and even there, ground the recommendation in what the codebase already does.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: thoughtbot
- Source: thoughtbot/rails-consultant
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.