ramp CLAUDE.md

ramp CLAUDE.md is an instructions file for Claude Code from gf-labs/ramp. It costs 3,819 tokens per session, scanned A, original, MIT.

Repository instructions for ramp, a learning tool for developers working with Claude Code, the codebase, and team workflows. They explain its design principles, project goals, and operating structure.

In plain words
What is it for?
Use them when prioritising or implementing work in the ramp project and its learning workflows.
Why use it?
They help an agent make changes that fit the project's focus on demonstrated ability rather than activity alone.

Instructions file for Claude Code

Written for Claude Code: PostToolUse hook event. Also seen: reads .claude/ paths; mentions CLAUDE.md; mentions subagents.

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 instructions/gf-labs/ramp/claude-md
Clone the repo
git clone --depth 1 https://github.com/gf-labs/ramp

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 ramp CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/gf-labs/ramp/claude-md.svg)](https://agentmods.dev/instructions/gf-labs/ramp/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/gf-labs/ramp/claude-md"><img src="https://agentmods.dev/badge/instructions/gf-labs/ramp/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 3,819 This file is loaded in full into every session.
When invoked 3,819 The same file — it is already loaded in full.
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 5.1 $0.03819 $0.03819
Opus 5 $0.01910 $0.01910
Sonnet 5 $0.00764 $0.00764
Haiku 4.5 $0.00382 $0.00382

Measured 5d ago against content hash c01928f658de, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

ramp 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 5d 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 · 190 lines

How it starts

The opening of the file, as written. The whole thing — 190 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Design philosophy

ramp measures demonstrated mastery ([✓]) over self-reported engagement ([~]) — capability growth, not activity metrics. The knowledge graph, spaced repetition, XP tiers, and Feynman-style verification all express that single thesis: a learning tool should track what you can do, not what you've clicked through.

It is also a working demonstration of Claude Code's extension model — built entirely from slash commands, hooks, an MCP server, a plugin manifest, and the settings hierarchy, with no application code.

When making prioritization decisions, ask: what best advances that capability-over-engagement thesis while showcasing Claude Code's extension model?

Project

ramp activates a learning mode — Claude becomes a co-pilot for ramping developers up on Claude Code as an organizational tool, the codebase, and the team's workflows (commits, PRs, testing, CI). It scans the environment, assesses the user, delivers a personalized knowledge graph and learning path, then stays engaged to work through exercises together. Writes an ONBOARDING.md artifact and updates ~/.claude/ramp/graphs/[topic].md.

Companies deploy it as a project-level command for engineer onboarding. Solo devs use it to level up.

Three operating modes

  • Mode A (empty repo): Claude asks what to build, then builds it while narrating every Claude Code capability it uses
  • Mode B (single repo): Scans codebase → assesses user → integrated learning path grounded in the actual repo
  • Mode C (multi-repo directory): Surveys all repos, user picks focus, then proceeds as Mode B

Structure

commands/up.md               # Adaptive onboarding command — manages knowledge graph
commands/calibrate.md        # Placement worksheet front door — writes .ramp/calibrate.md; claims seed the graph
commands/check.md            # Check back the active task — grade the worksheet, persist via save_graph, report the XP delta
commands/list.md             # Read-only topic catalog (CLI-backed) — what exists, where you've started
commands/help.md             # 60-second orientation — what ramp is + the command map
commands/tree.md             # Read-only knowledge graph viewer
commands/review.md           # Spaced repetition review command
commands/cheatsheet.md       # Personal reference: demonstrated skills + evidence trails
commands/pin.md              # Mid-session checkpoint — status, node save, optional snapshot
commands/ingest.md           # Generate topic schemas from external sources (PDF, URL, file)
commands/wrap.md             # End-of-session knowledge harvest — node upgrades, SR schedule, snapshot
.claude-plugin/plugin.json   # Plugin manifest (name, version, description)
.claude-plugin/marketplace.json # Marketplace catalog (source, category, install metadata)
hooks/hooks.json             # Plugin hook config (PostToolUse + SessionStart)
.github/workflows/test.yml   # CI — ruff lint + pytest matrix (3.8 floor + current) on develop/main + PRs
.github/workflows/release-gate.yml # CI — manifest ↔ CHANGELOG ↔ tag consistency for → main
.github/workflows/release.yml # CD — on a v* tag: lint+test gate, manifest↔tag assert, GitHub Release from the CHANGELOG section
.github/dependabot.yml       # Weekly SHA-pin bumps for GitHub Actions
.claude/settings.json        # Project hook config (file-size-warn + skill-observer)
topics/claude-code.md        # Claude Code meta-topic (sources 5 sub-topics, 81 nodes total)
topics/getting-started.md    # Getting started sub-topic (12 nodes)
topics/build.md              # Build sub-topic (32 nodes)
topics/configuration.md      # Configuration sub-topic (13 nodes)
topics/deployment.md         # Deployment sub-topic (11 nodes)
topics/administration.md     # Administration sub-topic (13 nodes)
topics/best-practices.md     # Best practices topic schema (15 nodes)
topics/mcp-development.md    # MCP development topic schema (29 nodes)
topics/anthropic-api.md      # Anthropic API topic schema (18 nodes)
topics/claude-code-internals.md # Empirically-verified Claude Code internals (5 nodes — undocumented behaviors)
topics/git.md                # Git version control topic schema (27 nodes — standalone, Pro Git-sourced; first breadth topic)
topics/bash.md               # Bash scripting topic schema (29 nodes — standalone, GNU Bash Reference Manual-sourced)
topics/python.md             # Python language topic schema (31 nodes — standalone, Python Tutorial + Language Reference-sourced)
scripts/skill-observer.py    # Passive observer hook (PostToolUse + SessionStart)
scripts/file-size-warn.py    # PostToolUse hook — warns when .md files exceed 600 lines
scripts/setup-mcp.py         # Provisions .venv + registers MCP server (opt-in — run manually, not on SessionStart)
mcp/server.py                # knowledge-graph MCP server (read/write graphs; swappable backend)
mcp/start.sh                 # MCP launch wrapper — .venv python; avoids macOS symlink trap
ramp_core.py                 # stdlib-only kernel — write side (XP · SR dates · validate · lock) + read side (catalog · summary · node_count · graph_nodes) + detect side (schema ## Probes → run_detection) + lint side (schema conformance — problems vs advisories); imported by skill-observer + mcp/server; CLI-backed for list/help/tree/detect/lint
tests/                       # stdlib pytest suite (ramp_core, xp, detection, schema-lint, normalization, symlinks, setup-mcp, file-size-warn) + server tests under .venv — importlib-loaded
requirements.txt             # MCP server deps — only to run mcp/server.py, not to test
requirements-dev.txt         # pytest (test suite) — installed by the matrix, incl. the 3.8 leg
requirements-lint.txt        # ruff pin — shared by test.yml's lint job and release.yml, kept out of the 3.8 leg
pyproject.toml               # Tooling config — ruff lint rules (target py38; no packaging, kernel is stdlib-only)
docs/tree-format.md          # Annotated v3 knowledge graph format example
docs/docs-map.md             # Maps all doc pages to topics and nodes
docs/topic-authoring.md      # Normative topic-authoring contract — skill definition, schema anatomy, probe/reference grammar, validation gate (lint-checked)
BACKLOG.md                   # Pointer to TaskWarrior backlog (task project:business.ramp)
.mcp.json.example            # MCP server config template (copy → .mcp.json, fill in paths)
README.md                    # Install instructions, modes, company deployment guide
LICENSE                      # MIT license
.gitignore                   # Ignores .venv/, .claude/commands/, docs/superpowers/, .mcp.json, .ramp/

Read the full file on GitHub · 190 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. 5d ago First seen · 190 lines · 3,819 tokens per session scan A c01928f658de

Subscribe to this mod's changes

ramp CLAUDE.md is an instructions file published in the GitHub repository gf-labs/ramp (2 stars, last pushed 1mo ago), licensed MIT. It adds 3,819 tokens to every session, about $0.0191 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.

Related

Other instructions, from other repositories

living-manual CLAUDE.md

Instructions for dougstanford/living-manual, covering living-manual, start of every session, how work lands — non-negotiable, user's manual (living-manual) and releasing.

dougstanford/living-manual · 1,108 tokens

ship-it open-pr.instructions.md

Instructions for khrichtchatyi/ship-it: When the user asks to open, create, or submit a pull request, follow the skill in skills/open-pr/SKILL.md in full: verify gh is authenticated, resolve the target repository and default branch, push the current branch, open one pull request with a title and body derived from the…

khrichtchatyi/ship-it · 84 tokens

agent-skills CLAUDE.md

Instructions for mindbox-cloud/agent-skills, covering agent-skills — developer guide for claude code, repository purpose, repository structure, how to add a new plugin and step 1 — create the directory structure.

mindbox-cloud/agent-skills · 1,143 tokens

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,104 tokens

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens