What is upv-mcp CLAUDE.md?

Project instructions for upv-mcp, a local server that provides access to the Polytechnic University of Valencia's academic calendar and PoliformaT system. They define its architecture, data-handling rules, commands, and tool conventions.

Use them when adding or changing sources, tools, server code, tests, configuration, or credentials in upv-mcp. They cover installation, tests, type checking, linting, and server startup.

They prevent code from crossing forbidden boundaries, committing real calendar data, or exposing credentials. They also define the checks that must pass before a change is finished.

How pgomsan/upv-mcp configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to pgomsan/upv-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.

Copy the file

curl -O https://raw.githubusercontent.com/pgomsan/upv-mcp/master/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/pgomsan/upv-mcp

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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/pgomsan/upv-mcp/claude-md/github.svg)](https://agentmods.dev/instructions/pgomsan/upv-mcp/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/pgomsan/upv-mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/pgomsan/upv-mcp/claude-md/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.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/pgomsan/upv-mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/pgomsan/upv-mcp/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 875 This file is loaded in full into every session.

When invoked 875 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.00875	$0.00875
Opus 5.5	$0.00350	$0.00350
Sonnet 5.5	$0.00175	$0.00175
Haiku 4.5	$0.00088	$0.00088

Measured 1mo ago against content hash 10a5ffec4eb8, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

upv-mcp CLAUDE.md 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.

CLAUDE.md · 69 lines

How it starts

The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.

upv-mcp

Servidor MCP local con el calendario academico de la UPV y PoliformaT. SDK mcp 2.x (la clase es MCPServer; en la 1.x era FastMCP). Transporte stdio.

Comandos

uv sync                  # instalar
uv run pytest -q         # tests
uv run mypy              # strict, debe salir limpio
uv run ruff check .      # lint (ruff format . para formatear)
uv run upv-mcp           # arrancar el servidor (stdio)
uv run upv-mcp-config show   # ver que credenciales estan configuradas

Los cuatro primeros deben pasar antes de dar por terminado un cambio.

Invariantes de arquitectura

Romperlos es el unico motivo real para rechazar un PR aqui.

sources/ no importa mcp. Solo red + parseo -> modelos de dominio.
tools/ no importa mcp, ni httpx2, ni icalendar. Solo modelos + repository.
server.py es el UNICO fichero que importa mcp.
tests/test_architecture.py lo comprueba con ast: rompe la build si se viola.

Datos

El .ics real no se commitea. Los fixtures de tests/fixtures/ son anonimizados y conservan la estructura exacta del generador de la UPV.
Credenciales (URL iCal con token, usuario y clave UPV) van al llavero via upv-mcp-config. Nunca a un fichero ni a un .env.
Todo datetime se normaliza a Europe/Madrid al parsear, nunca en tools/.
Peculiaridades de cada origen: docstrings de sources/ics.py, poliformat.py y cas.py. Leelos antes de tocarlos; no estan documentadas en ningun sitio publico.
PoliformaT NO permite login por API: solo CAS. Un login rechazado no se reintenta nunca, porque bloquea la cuenta.
Los examenes vienen de su propio iCal (opcional). Nunca los deduzcas del titulo de una tarea: hay tareas llamadas "Examen" que no lo son, y al reves.

Convenciones de tools

Nombre en snake_case, verbo primero: get_schedule, list_upcoming_deadlines.
Una tool por modulo en tools/, registrada en server.py.
Navegar contenido va como resource; descargarlo y convertirlo, como tool.
La descripcion es codigo de produccion, no documentacion: decide si el modelo elige bien. Di que devuelve, cuando usarla y cuando NO, redirigiendo por nombre a la tool correcta.
Devolver datos estructurados y compactos: filtrar, ordenar y truncar antes de responder. Al truncar, meta.truncated = true.
Si una tool no puede ver algo, decirlo en meta.coverage_note. Es lo que impide que el modelo afirme "no tienes examenes" cuando no los esta mirando.

Read the full file on GitHub · 69 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.

1mo ago Changed · +4 lines · +84 tokens per session 10a5ffec4eb8
2mo ago First seen · 65 lines · 791 tokens per session scan A 505403f17237

Subscribe to this mod's changes

upv-mcp CLAUDE.md is an instructions file published in the GitHub repository pgomsan/upv-mcp (1 stars, last pushed 1mo ago), licensed MIT. It adds 875 tokens to every session, about $0.0035 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.

Learning and teaching

upv-mcp CLAUDE.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
upv-mcp CLAUDE.mdpgomsan/upv-mcp	Instructions file	1	1mo ago	MIT	A	875	Claude Code
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	yesterday	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	yesterday	MIT	A	4,680	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	5,516	GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	274k	2d ago	MIT	A	4,313	Codex, OpenCode

upv-mcp: Instructions file for Claude Code