Borrowing it
Nothing to install: this file belongs to vargahis/monarch-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/vargahis/monarch-mcp/main/.claude/skills/test-monarch-mcp/SKILL.mdgit clone --depth 1 https://github.com/vargahis/monarch-mcpWrote 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/vargahis/monarch-mcp/test-monarch-mcp)<a href="https://agentmods.dev/skills/vargahis/monarch-mcp/test-monarch-mcp"><img src="https://agentmods.dev/badge/skills/vargahis/monarch-mcp/test-monarch-mcp/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/skills/vargahis/monarch-mcp/test-monarch-mcp"><img src="https://agentmods.dev/badge/skills/vargahis/monarch-mcp/test-monarch-mcp.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 | $0.00086 | $0.05713 |
| Opus 5 | $0.00043 | $0.02857 |
| Sonnet 5 | $0.00017 | $0.01143 |
| Haiku 4.5 | $0.00009 | $0.00571 |
Grade A, and why
test-monarch-mcp 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 12d 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 — 522 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Test Monarch MCP Skill
You are the orchestrator of a comprehensive test suite for the Monarch Money MCP server.
You do not call Monarch tools yourself. Instead you dispatch a subagent per phase (via the Task tool); each subagent runs that phase's tests, asserts on shape/key fields only, and returns a compact PASS/FAIL summary. You aggregate the summaries, track created resources, drive cleanup, and print the final report. This keeps large tool payloads (transaction lists, rule lists, full objects) inside the subagents and out of your context.
Run tests across 14 phases, track results, and clean up after yourself.
Scope of this skill: it validates that the agent uses the MCP tools correctly — right tool, right params, correct interpretation of responses (happy paths, agent-judgment cases, and one representative graceful-error case per tool family). It does not stress-test tool robustness. Adversarial/edge inputs and live-API error paths are covered by the deterministic e2e suite in
tests/integration/(run withMONARCH_LIVE_TESTS=1 pytest tests/integration -m integration).
Mode Support
This test suite supports two modes, auto-detected at startup:
- Read-only mode (default): Tests 27 read-only tools (58 tests). Write-dependent tests are skipped. No data is created, modified, or deleted.
- Write-enabled mode (
--enable-write): Tests all 44 tools (110 tests). Creates, modifies, and deletes data on your live Monarch account. Self-cleaning.
Execution Model
Your job (orchestrator)
- Detect the server mode and get user confirmation (see Pre-flight).
- Dispatch the discovery subagent, then one subagent per phase, then (write mode) the cleanup subagent.
- Parse each subagent's compact JSON return. Never request or echo a raw tool payload.
- Maintain
mcp-test-state.json; merge the created-resource IDs each subagent reports. - Print the final summary.
Dispatching a subagent
Use the Task tool (subagent type: general-purpose). The prompt you pass must contain:
What ships with it
14 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.
- references/account-management.md 2.2 KB
- references/accounts-and-holdings.md 2.3 KB
- references/analytics-tools.md 1000 B
- references/auth-tools.md 1.4 KB
- references/budgets-and-cashflow.md 3.6 KB
- references/categories.md 2.2 KB
- references/read-only-tools.md 2.6 KB
- references/recurring-merchant.md 4.6 KB
- references/tag-crud.md 2.9 KB
- references/transaction-crud.md 7.2 KB
- references/transaction-details-and-splits.md 2.1 KB
- references/transaction-rules.md 6.1 KB
- references/transaction-tagging.md 2.2 KB
- references/transactions-read.md 3.5 KB
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.
- 12d ago First seen · 522 lines · 86 tokens per session scan A 80887d5ed52a
test-monarch-mcp is a skill published in the GitHub repository vargahis/monarch-mcp (12 stars, last pushed 3mo ago), licensed MIT. It adds 86 tokens to every session and 5,713 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. It comes from a forked repository.
Other skills, from other repositories
tdd-methodology-expert
Use proactively when you need to implement features or fix bugs using strict Test-Driven Development (TDD) methodology. This agent should be activated for any coding task that requires writing new functionality, refactoring existing code, or ensuring comprehensive test coverage, but should not be used for any…
laravel-dusk
Laravel Dusk - Browser automation and testing API for Laravel applications. Use when writing browser tests, automating UI testing, testing JavaScript interactions, or implementing end-to-end tests in Laravel.
qa-test-planner
Generate comprehensive test plans, manual test cases, regression test suites, and bug reports for QA engineers. Includes Figma MCP integration for design validation.
langsmith-observability
LLM observability platform for tracing, evaluation, and monitoring. Use when debugging LLM applications, evaluating model outputs against datasets, monitoring production systems, or building systematic testing pipelines for AI applications.
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
agent-evaluation
Testing and benchmarking LLM agents including behavioral testing, capability assessment, reliability metrics, and production monitoring—where even top agents achieve less than 50% on real-world benchmarks Use when: agent testing, agent evaluation, benchmark agents, agent reliability, test agent.