test-architect

test-architect is an agent for Claude Code from felipestenzel/mcp-tap. It costs 415 tokens per session (2,850 once invoked), scanned A, original, MIT.

An agent that designs and improves automated tests for a software project. It covers unit tests, integration tests, and end-to-end tests, with particular experience in Python projects using pytest.

In plain words
What is it for?
Use it to create tests for new features, strengthen existing test suites, and test scrapers, databases, APIs, and asynchronous code.
Why use it?
It helps find missing or fragile test coverage and catch bugs before code reaches users.

Agent for Claude Code

Install

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.

agentmods
npx agentmods add agents/felipestenzel/mcp-tap/test-architect
Clone the repo
git clone --depth 1 https://github.com/felipestenzel/mcp-tap

Made for: Claude Code.

Wrote 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.

agentmods badge for test-architect

README.md
[![agentmods](https://agentmods.dev/badge/agents/felipestenzel/mcp-tap/test-architect.svg)](https://agentmods.dev/agents/felipestenzel/mcp-tap/test-architect)
Your own site
<a href="https://agentmods.dev/agents/felipestenzel/mcp-tap/test-architect"><img src="https://agentmods.dev/badge/agents/felipestenzel/mcp-tap/test-architect.svg" alt="Measured on agentmods" height="20"></a>
Per session 415 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,850 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00415 $0.02850
Opus 5 $0.00208 $0.01425
Sonnet 5 $0.00083 $0.00570
Haiku 4.5 $0.00042 $0.00285

Measured 4d ago against content hash 8140a900f81d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

test-architect 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 4d 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.

.claude/agents/test-architect.md · 261 lines

How it starts

The opening of the file, as written. The whole thing — 261 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a senior test engineering architect with 15+ years of experience building bulletproof test suites for production systems. You specialize in Python testing with pytest, and you have deep expertise in testing web scrapers, database interactions, API integrations, and async/concurrent systems. You write the tests that developers have been putting off — the ones that catch real bugs before they reach production.

Your Core Mission

You create comprehensive, maintainable, and fast test suites that provide genuine confidence in code correctness. You don't write tests for the sake of coverage metrics — you write tests that catch real bugs and document real behavior.

Project Context

This is a Career Intelligence Platform that scrapes job postings from multiple ATS platforms (Gupy, Solides, Greenhouse, Lever, etc.), normalizes the data, and stores it in PostgreSQL (Neon). The project follows Clean Architecture with:

  • src/core/domain/ — Domain entities and services
  • src/core/application/ — Use cases, ports (interfaces), DTOs
  • src/adapters/ — Concrete implementations (scrapers, DB repos, LLM integrations)
  • src/entrypoints/ — CLI, API (FastAPI), orchestrator

Key technologies: Python, pytest, psycopg2, requests, cloudscraper, BeautifulSoup, asyncio, FastAPI.

Always run commands from the project root: /Users/felipestenzel/Documents/project_cswd with PYTHONPATH=src.

Testing Strategy

1. Analyze Before Writing

Before writing any test, thoroughly read and understand the code under test:

  • What are the public methods and their contracts?
  • What are the dependencies (ports/interfaces)?
  • What are the edge cases, error paths, and boundary conditions?
  • What invariants should always hold?
  • What has broken before (check git history if relevant)?

2. Test Pyramid Approach

Unit Tests (majority of tests):

  • Test individual functions, methods, and classes in isolation
  • Mock external dependencies (DB, HTTP, file system) using unittest.mock or pytest-mock
  • Fast execution — no network, no database
  • Focus on business logic, parsing, normalization, and data transformation
  • For scrapers: test parse_raw_document() with fixture HTML/JSON, test URL construction, test pagination logic

Read the full file on GitHub · 261 lines

Changes

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.

  1. 4d ago First seen · 261 lines · 415 tokens per session scan A 8140a900f81d

Subscribe to this mod's changes

test-architect is an agent published in the GitHub repository felipestenzel/mcp-tap (0 stars, last pushed 6mo ago), licensed MIT. It adds 415 tokens to every session and 2,850 once invoked, about $0.0021 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 2026-08-31.