Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/thoughtbot/rails-consultant/feature-devnpx skills add thoughtbot/rails-consultant --skill feature-devgit clone --depth 1 https://github.com/thoughtbot/rails-consultantWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/thoughtbot/rails-consultant/feature-dev)<a href="https://agentmods.dev/skills/thoughtbot/rails-consultant/feature-dev"><img src="https://agentmods.dev/badge/skills/thoughtbot/rails-consultant/feature-dev.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00083 | $0.03911 |
| Opus 5 | $0.00042 | $0.01956 |
| Sonnet 5 | $0.00017 | $0.00782 |
| Haiku 4.5 | $0.00008 | $0.00391 |
Grade A, and why
feature-dev scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 170 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Feature Development
You are helping a developer take one feature slice from a rough idea to reviewed, committed code. This is not autonomous coding — it's a guided workflow that borrows its backbone from systematic feature development but threads four disciplines through it: slicing to define the work, TDD to build it, code review to harden it, and a clean commit to land it.
The workflow's whole reason for existing is to resist the pull toward premature code. Each phase earns the right to the next: you don't slice until you understand the feature, you don't build until the slice is sharp, you don't consider it done until the diff has been reviewed, and you don't commit until it's green. Hold that line — the value is in the sequence, not any single step.
Scope discipline: this workflow refines and builds exactly one slice. If the work is actually an epic — several independently shippable pieces — you'll surface that during framing and help the user pick the single slice to build now. Building more than one slice in a pass defeats the point; the payoff is a tight loop of shape → build → review on a small, real increment.
Leanness discipline: a small slice does not guarantee a small diff. Slicing controls what ships; it does nothing to stop the implementation from bloating with speculative abstractions, options nothing uses yet, defensive branches no test demands, or gold-plating. Those are a separate failure mode, and this workflow fights them separately. The target is a production-code diff under ~300 lines (excluding comments and blank lines; test code does not count against this ceiling, though bloated tests are their own smell). Treat that number as a design constraint you carry from Phase 3 onward, not a gate you discover at the end.
Use TodoWrite to track the phases so the user can see where they are.
Phase 1: Frame the work
Goal: understand what the user wants to build, and confirm it's a single slice before investing in anything else.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 170 lines · 83 tokens per session scan A a9d394ff2ffd
feature-dev is a skill published in the GitHub repository thoughtbot/rails-consultant (23 stars, last pushed 1mo ago), licensed MIT. It adds 83 tokens to every session and 3,911 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
upgrade-cleanup
Clean up after (or abandon) a Rails upgrade. Drop NextRails.next? and NextRails.current? branches and retire dual-boot scaffolding (Gemfile.next, Gemfile.next.lock, conditional Gemfile groups), keeping either the next or the current version. Trigger when the user says they are done with the upgrade, want to clean up…
recursive-decomposition
Handle tasks that exceed the context window by decomposing them: size and filter the input, chunk it, run recursive sub-agents on independent parts, verify on small windows, and synthesise programmatically, following the Recursive Language Models (RLM) research by Zhang, Kraska and Khattab (2025). Use when a task…
phase4-spec
The Mileva Method (CRISP) — Phase 4: Spec. Full implementation readiness package including solution design, UX direction, tech stack, backlog, MVP prioritization, risk assessment, and AI architecture for Claude Code projects — CLAUDE.md, skill mapping, AI specs, sprint planning, quality gates. Triggers on "spec"…
swarm
Run a multi-agent audit of a codebase by spawning specialized parallel subagents (security, performance, tests, architecture, dead-code), then synthesize their findings into a single prioritized action plan. Use this whenever the user runs /swarm, asks to "audit the repo," "review this codebase," "find issues across…
crisp-execute
CRISP Execute — Sprint execution loop with change request management, security gates, product gates, and stakeholder reporting. Use after Phase S (Spec) is complete and crisp-state.json shows readyforexecute: true. Triggers on "start sprint", "begin build", "execute", "sprint 1", "run sprints", or when handed off from…
phase1-clarify
The Mileva Method (CRISP) — Phase 1: Clarify. Problem definition, eliciting, painkiller test, buy vs build, AI justification gate, Go/No-Go. Use at the start of any AI implementation project. Triggers on "clarify", "phase 1", "start discovery", "define the problem", "what problem are we solving", or at the beginning…