agentmods.dev

oncofiles: Instructions file for Claude Code

CLAUDE.md

oncofiles CLAUDE.md is an instructions file for Claude Code from peter-fusek/oncofiles. It costs 17,163 tokens per session, scanned A, original, MIT.

A project guide for oncofiles, a document-management service for cancer patients and caregivers. It connects Google Drive, Gmail, and Calendar to AI assistants and describes the project's setup, file naming, and development commands.

In plain words
What is it for?
Use it when developing or testing oncofiles, running its MCP server, managing migrations, checking documents, or following its required filename and Python conventions.
Why use it?
It gives coding agents the project's structure and conventions before they change code. This helps preserve document names, compatibility, tests, and formatting rules.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: mentions Claude Code.

This is peter-fusek/oncofiles's own configuration. It tells Claude Code how to work on oncofiles itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything oncofiles configures →

not rated 5repo changed today A scan 17,163 tokens original MIT
View source ↗ peter-fusek/oncofiles oncofiles.com
Reuse

Borrowing it

Nothing to install: this file belongs to peter-fusek/oncofiles. 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/peter-fusek/oncofiles/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/peter-fusek/oncofiles

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 oncofiles CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/peter-fusek/oncofiles/claude-md/github.svg)](https://agentmods.dev/instructions/peter-fusek/oncofiles/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/peter-fusek/oncofiles/claude-md"><img src="https://agentmods.dev/badge/instructions/peter-fusek/oncofiles/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.

agentmods 80×15 button for oncofiles CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/peter-fusek/oncofiles/claude-md"><img src="https://agentmods.dev/badge/instructions/peter-fusek/oncofiles/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 17,163 This file is loaded in full into every session.
When invoked 17,163 The same file — it is already loaded in full.
Security scan A 1 finding. 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.17163 $0.17163
Opus 5.5 $0.06865 $0.06865
Sonnet 5.5 $0.03433 $0.03433
Haiku 4.5 $0.01716 $0.01716

Measured today against content hash 889b1656b133, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

oncofiles CLAUDE.md scanned grade A with 1 finding 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 today.

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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- **Session cookie URL-encoding (#495)**: Python `SimpleCookie` treats `@` (present in every email address) as outside its `_LegalValueChars` and wraps the entire cookie value in double-quotes. Browsers echo those litera
CLAUDE.md · 148 lines

How it starts

The opening of the file, as written. The whole thing — 148 lines — stays where its author put it; the contents beside it link to each section on GitHub.

oncofiles

AI-powered medical document management for cancer patients and caregivers. Connects Google Drive, Gmail, and Calendar to Claude, ChatGPT, and any MCP client.

Quick start

uv sync --extra dev
uv run pytest
uv run ruff check

Project structure

  • src/oncofiles/ — main package
  • tests/ — pytest tests
  • migrations/ — SQL schema migrations
  • data/ — local dev SQLite database (gitignored)

Conventions

  • Python 3.12+, async-first
  • FastMCP 3.0+ for MCP server
  • Pydantic for data models
  • ruff for linting/formatting
  • Filenames follow standard convention: YYYYMMDD_PatientName_Institution_Category_DescriptionEN.ext
    • Separators: underscores only
    • Description: English, CamelCase, max 60 chars
    • Category tokens: Labs, Report, Pathology, CT, USG, Genetics, Surgery, SurgicalReport, Prescription, Referral, Discharge, DischargeSummary, ChemoSheet, Reference, Advocate, Other, Vaccination, Dental, Preventive
    • Legacy formats (space+dash, underscore-separated) still parsed for backward compat

Key commands

  • uv run oncofiles-mcp — run MCP server (stdio)
  • MCP_BEARER_TOKEN=test MCP_TRANSPORT=streamable-http uv run oncofiles-mcp — run HTTP server locally
  • uv run pytest — run tests
  • uv run ruff check --fix — lint and auto-fix
  • uv run ruff format — format code

Deployment

  • Self-hosted Mac mini (since 2026-11-06): Coolify in a Colima Linux VM, Traefik + Let's Encrypt in front, oncofiles.com streamable-http at /mcp. Build pack Railpack, install uv sync, start uv run oncofiles-mcp. Forpsi and Railway are retired. Host/network details live in the private RUNBOOK, never in this public repo (no IPs, IDs, tokens).
  • Push to main deploys via the CI deploy job (.github/workflows/ci.yml): after lint+test pass, the runner joins the tailnet as an ephemeral tag:ci node and POSTs to the Coolify API, then smoke-tests /health. Kill switch: repo variable DEPLOY_ENABLED. Always run /deploy-check after a push.
  • Env-only changes are cheap: with an unchanged commit SHA, Coolify logs No build configuration changed & image found … Build step skipped and just recreates the container.
  • Never tick "Available at Buildtime" on a new env var unless the build genuinely needs it — Railpack writes buildtime vars into a .env that mise/aqua read (#550).
  • DB is a local SQLite file (aiosqlite, WAL) on the Coolify volume, path DATABASE_PATH. There is no cloud DB option: any non-empty TURSO_* env var makes startup fail (server._assert_local_database, #617). Data survives rebuilds; a failed deploy keeps the old container running.
  • Google: a dedicated GCP project, named in the RUNBOOK (OAuth client, External / In production / unverified, 100-user cap accepted). GOOGLE_INTEGRATION_ENABLED stays false until the curated-folder pipeline (#562) ships.
  • CI is lint + test + deploy.

Read the full file on GitHub · 148 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. today Changed · +5 lines · +1,831 tokens per session 889b1656b133
  2. yesterday Changed · +15 lines · +4,262 tokens per session bd25ce72d566
  3. 2d ago Changed · +118 tokens per session efb56d5ff3a0
  4. 3d ago Changed · +4 lines · +838 tokens per session ef68edc2708d
  5. 4d ago Changed · +6 lines · +1,176 tokens per session b60a35e8d2a5
  6. 6d ago Changed · +30 tokens per session 6f38ad47b6a6
  7. 2mo ago First seen · 118 lines · 8,908 tokens per session scan A b25980ccf38a

Subscribe to this mod's changes

oncofiles CLAUDE.md is an instructions file published in the GitHub repository peter-fusek/oncofiles (5 stars, last pushed today), licensed MIT. It adds 17,163 tokens to every session, about $0.0687 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

DocumentsPythonfastmcphealthmcpmcp-servermedical
Related

oncofiles CLAUDE.md compared with instructions from other repositories

oncofiles CLAUDE.md next to 6 instructions from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
oncofiles CLAUDE.mdpeter-fusek/oncofilesInstructions file5todayMITA17,163Claude Code
vscode buildNext.instructions.mdmicrosoft/vscodeInstructions file195ktodayMITA9,719GitHub Copilot
next.js AGENTS.mdvercel/next.jsInstructions file143ktodayMITA7,461Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscodeInstructions file195ktodayMITA5,685GitHub Copilot
langchain AGENTS.mdlangchain-ai/langchainInstructions file149ktodayMITA4,680Codex, OpenCode
vscode copilot-instructions.mdmicrosoft/vscodeInstructions file195ktodayMITA3,086GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harnessInstructions file280k2d agoMITA4,325Codex, OpenCode