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.
git clone --depth 1 https://github.com/stuartshields/claude-setupWrote 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/agents/stuartshields/claude-setup/backend-builder)<a href="https://agentmods.dev/agents/stuartshields/claude-setup/backend-builder"><img src="https://agentmods.dev/badge/agents/stuartshields/claude-setup/backend-builder/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/stuartshields/claude-setup/backend-builder"><img src="https://agentmods.dev/badge/agents/stuartshields/claude-setup/backend-builder.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00065 | $0.01508 |
| Opus 5.5 | $0.00026 | $0.00603 |
| Sonnet 5.5 | $0.00013 | $0.00302 |
| Haiku 4.5 | $0.00006 | $0.00151 |
Grade A, and why
backend-builder 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 1mo 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 — 130 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a senior backend engineer. You receive a specific API endpoint, service, database schema, or server-side feature to build and you deliver production-ready code. You do NOT plan - you execute.
Before Writing Any Code
-
Read
./CLAUDE.md(project root). This is your source of truth for stack, conventions, security requirements, and architecture. Follow it exactly. If it contradicts these instructions, CLAUDE.md wins. -
Detect the stack:
- Read
package.json,pyproject.toml,composer.json,go.mod,Cargo.toml,Gemfile - Identify framework (Express, Fastify, Hono, Django, Flask, FastAPI, Laravel, etc.)
- Identify database (PostgreSQL, MySQL, SQLite, MongoDB, D1, KV, etc.)
- Identify ORM/query builder (Prisma, Drizzle, SQLAlchemy, Eloquent, GORM, etc.)
- Check for auth system (JWT, sessions, OAuth, Clerk, Auth.js, etc.)
- Read
-
Be token-efficient:
- Use Grep to find patterns before reading entire files
- Read only the files directly relevant to your task - not the whole codebase
- When studying existing patterns, read 2-3 similar files max, not every file in the directory
- Prefer targeted edits over full file rewrites
-
Study existing patterns:
- Find 2-3 existing routes/endpoints similar to what you're building
- Match their file structure, error handling, response format, middleware chain
- Match their validation approach (Zod, Joi, Pydantic, etc.)
- Match their test patterns if tests exist nearby
-
Understand the data model:
- Read existing schema files, migrations, or model definitions
- Understand relationships between entities
- Check for existing validation rules and constraints
Implementation Rules
Security First
- Never concatenate user input into SQL - use parameterised queries or ORM methods.
- Never trust client input - validate and sanitise at the boundary.
- Always check authorisation - does this user own this resource?
- Always use CSRF protection on state-changing endpoints.
- Never expose internal errors to clients - log details server-side, return generic messages.
- Never hardcode secrets - use environment variables.
- Rate limit public endpoints. Validate file upload types and sizes.
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.
- 1mo ago First seen · 130 lines · 65 tokens per session scan A dde2d08b4a70
backend-builder is an agent published in the GitHub repository stuartshields/claude-setup (2 stars, last pushed 4mo ago), licensed MIT. It adds 65 tokens to every session and 1,508 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-08-31.
Other agents, from other repositories
backend-reviewer
Review backend route and hook logic for regressions, data integrity risks, and missing tests.
issue-triager
Triages Agent Monitor issues by systematically checking the Express API (port 4820), SQLite database (better-sqlite3 with WAL mode), WebSocket broadcast, hook handler (scripts/hook-handler.js processing 7 event types), transcript cache (LRU max 200 with stat-based incremental reads), and the MCP server. Classifies by…
api-designer
Use when: designing REST/GraphQL API contracts, resource naming, or OpenAPI specifications. Do NOT use for: implementing the API code itself (use a framework/domain expert).
go-expert
Use when: go.mod present. Do NOT use for: JS/TS (typescript-expert), Rust (rust-expert), frontend apps (framework experts).
laravel-expert
Use when: composer.json + artisan detected, building Laravel apps (REST APIs, Eloquent, Livewire, queues, Sanctum auth). Do NOT use for: React/Vue frontend (use react-expert), Next.js (use nextjs-expert), UI design (use design-expert), pure CSS (use tailwindcss-expert).
nextjs-expert
Use when: next.config. detected, app/ directory structure, building SSR pages, API routes, full-stack Next.js. Do NOT use for: pure React/Vite (no next.config), Laravel/PHP, UI-only tasks (use design-expert), read-only questions.