claude-code-karma CLAUDE.md

Project instructions for Claude Code Karma, an application that reads Claude Code's local session files and displays them in a web dashboard.

In plain words
What is it for?
Use them when setting up, navigating, or changing the dashboard, API, frontend, or session-monitoring features.
Why use it?
They explain how the Python API and frontend are started and clarify that historical viewing is read-only while live session tracking is a separate setup level.

Instructions file

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/jayantdevkar/claude-code-karma/claude-md
Clone the repo
git clone --depth 1 https://github.com/JayantDevkar/claude-code-karma
Per session 2,168 This file is loaded in full into every session.
When invoked 2,168 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 $0.02168 $0.02168
Opus 5 $0.01084 $0.01084
Sonnet 5 $0.00434 $0.00434
Haiku 4.5 $0.00217 $0.00217

Measured 2d ago against content hash 5f2dea2e69ee, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

claude-code-karma 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 2d 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 · 262 lines

How it starts

The opening of the file, as written. The whole thing — 262 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.

Project Overview

Claude Code Karma is a full-stack application for monitoring and analyzing Claude Code sessions. It parses Claude Code's local storage (~/.claude/) and visualizes session data through a web dashboard.

Quick Start

# Clone (no submodule init needed — this is a monorepo)
git clone https://github.com/JayantDevkar/claude-code-karma.git
cd claude-code-karma

# Start API (terminal 1)
cd api
pip install -e ".[dev]" && pip install -r requirements.txt
uvicorn main:app --reload --port 8020

# Start frontend (terminal 2)
cd frontend
npm install && npm run dev

Open http://localhost:5180 to view the dashboard.

This is Tier 1 only — a read-only historical viewer. Karma's core feature is live session tracking (Tier 2 in SETUP.md); install through Tier 2 by default unless the user explicitly wants only historical browsing.

Repository Structure

This is a monorepo with all components in a single repository:

claude-code-karma/
├── api/                    # FastAPI backend (Python) - port 8020
├── frontend/               # SvelteKit frontend (Svelte 5) - port 5180
├── captain-hook/           # Claude Code hooks Pydantic library
└── hooks/                  # Hook scripts (symlinked from ~/.claude/hooks/)
    ├── live_session_tracker.py    # Tracks live session state across hook events
    ├── session_title_generator.py # Generates session titles via Haiku on SessionEnd
    └── plan_approval.py           # Handles ExitPlanMode permission requests

Each directory has its own CLAUDE.md with module-specific guidance.

Commands

API (Python/FastAPI)

cd api

# Run server
uvicorn main:app --reload --port 8020

# Run all tests
pytest

# Run specific test file
pytest tests/test_session.py -v

# Run API endpoint tests
pytest tests/api/ -v

# Run with coverage
pytest --cov=models --cov=routers

# Lint & format
ruff check models/ tests/ routers/
ruff format models/ tests/ routers/

Read the full file on GitHub · 262 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. 2d ago First seen · 262 lines · 2,168 tokens per session scan A 5f2dea2e69ee

Subscribe to this mod's changes

claude-code-karma CLAUDE.md is an instructions file published in the GitHub repository JayantDevkar/claude-code-karma (323 stars, last pushed 3d ago), licensed Apache-2.0. It adds 2,168 tokens to every session, about $0.0108 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.