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 instructions/aaronsb/texflow-mcp/claude-mdgit clone --depth 1 https://github.com/aaronsb/texflow-mcpWhat 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 | $0.00770 | $0.00770 |
| Opus 5 | $0.00385 | $0.00385 |
| Sonnet 5 | $0.00154 | $0.00154 |
| Haiku 4.5 | $0.00077 | $0.00077 |
Grade A, and why
texflow-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 yesterday.
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 — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TeXFlow MCP Server
Project Overview
TeXFlow is a LaTeX document compiler with an MCP interface. The AI operates on a structured document model (sections, paragraphs, figures, tables) and the server handles all LaTeX mechanics — packages, preamble, compilation.
Two entry flows:
- Scaffold:
document(action="create")→ empty document skeleton → AI builds content viaedittool - Markdown ingest:
document(action="ingest", source="file.md")→ populated model → AI refines layout
Both converge on an in-memory Document model that auto-saves to disk. LaTeX is only ever an output artifact.
Architecture
server.py # FastMCP entry point, 5 tool registrations
texflow/
model.py # Document model dataclasses
serializer.py # Model → .tex generation
ingestion.py # Markdown → model (mistune AST)
compiler.py # .tex → PDF (xelatex subprocess)
tools/
state.py # Shared session state, auto-save
document.py # create, ingest, outline, read
layout.py # configure typesetting
edit.py # insert, replace, delete, move blocks
render.py # compile, preview, tex export
reference.py # LaTeX documentation search
data/
font_map.json # Font name → LaTeX package
class_defaults.json # Document class defaults
latex_reference/ # Commands, symbols, packages, errors
tests/
test_model.py
test_serializer.py
test_ingestion.py
test_compiler.py
test_tools.py
fixtures/sample.md
5 MCP Tools
- document — create, ingest, outline, read
- layout — columns, fonts, paper, margins, headers/footers, TOC
- edit — insert/replace/delete/move blocks (section, paragraph, figure, table, code, equation, list, raw)
- render — compile to PDF, preview page as PNG, export .tex
- reference — search commands/symbols, package info, error help, style check, examples
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.
- yesterday First seen · 90 lines · 770 tokens per session scan A efbdf2f8127f
texflow-mcp CLAUDE.md is an instructions file published in the GitHub repository aaronsb/texflow-mcp (20 stars, last pushed 6mo ago), licensed MIT. It adds 770 tokens to every session, about $0.0038 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 instructions, from other repositories
TFG-TFM_EPS AGENTS.md
Instructions for jmrplens/TFG-TFM_EPS, covering agents.md — instrucciones para agentes de ia, contexto del proyecto, reglas de edición, archivos que el agente puede editar libremente and regla crítica de idioma.
TFG-TFM_EPS CLAUDE.md
Instructions for jmrplens/TFG-TFM_EPS, covering claude.md — instrucciones para claude, comandos de compilación, estructura de archivos, configuración del documento and claves principales.
TFG-TFM_EPS copilot-instructions.md
Instructions for jmrplens/TFG-TFM_EPS, covering reglas críticas, estructura del proyecto, configuración del documento, titulaciones disponibles and regla de idioma.
prfaq CLAUDE.md
Instructions for punt-labs/prfaq, covering how i write code, mandatory reading, no "pre-existing" excuse, standards and development workflow.
latexml-oxide CLAUDE.md
Instructions for dginev/latexml-oxide, covering claude.md, active priorities: faithful parity + beyond-perl arxiv runs, project overview, workspace structure and internal documentation.
z-spec CLAUDE.md
Instructions for punt-labs/z-spec, covering z-spec, mandatory reading, read this first, architecture and key architectural boundary: commands vs. surfaces.