TeleFuser CLAUDE.md

TeleFuser CLAUDE.md is an instructions file for coding agents from Tele-AI/TeleFuser. It costs 2,169 tokens per session, scanned A, original, Apache-2.0.

Project instructions for TeleFuser, a Python and PyTorch framework for running multimodal AI models with components such as CUDA, FastAPI, and Ray.

In plain words
What is it for?
Navigating, debugging, extending, and testing TeleFuser’s model pipelines, operations, kernels, scheduling, caching, and distributed processing.
Why use it?
They point an agent to the project’s source of truth, directory structure, documentation, tests, and development rules before it changes code.

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/tele-ai/telefuser/claude-md
Clone the repo
git clone --depth 1 https://github.com/Tele-AI/TeleFuser

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 TeleFuser CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/tele-ai/telefuser/claude-md.svg)](https://agentmods.dev/instructions/tele-ai/telefuser/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/tele-ai/telefuser/claude-md"><img src="https://agentmods.dev/badge/instructions/tele-ai/telefuser/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 2,169 This file is loaded in full into every session.
When invoked 2,169 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 $0.02169 $0.02169
Opus 5 $0.01085 $0.01085
Sonnet 5 $0.00434 $0.00434
Haiku 4.5 $0.00217 $0.00217

Measured 5d ago against content hash 3001dfb38743, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

TeleFuser 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 · 180 lines

How it starts

The opening of the file, as written. The whole thing — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.

TeleFuser - Agent Guidelines

Scope And Sources Of Truth

TeleFuser is a high-performance multimodal inference framework built with Python, PyTorch, CUDA, FastAPI, and Ray.

  • Treat the current repository, tests, and documentation as the API source of truth.
  • Use README.md for project structure, supported models, and documentation discovery.
  • Use CONTRIBUTING.md for development setup, contribution workflow, and coding standards.
  • Read the relevant guides under docs/en/ or docs/zh/ before changing a subsystem.
  • Prefer the closest maintained implementation and its tests over adding a new pattern.

Repository Map

telefuser/
  core/             Base abstractions, configuration, and module management
  pipelines/        Model-specific pipeline and stage implementations
  models/           DiT, VAE, text encoder, and other model architectures
  ops/              Public compile-aware operation dispatch
  kernel/triton/    Internal Triton kernel implementations
  schedulers/       Diffusion schedulers
  distributed/      FSDP, tensor/pipeline/sequence parallelism, and communication
  feature_cache/    Feature caching implementations
  cache/            General cache management
  offload/          CPU and device offload strategies
  metrics/          Runtime measurement and raw metrics
  orchestrator/     Request and actor-based streaming orchestration
  worker/           Distributed worker management
  service/          FastAPI and LiveKit-backed services
  client/           Python client SDK
  entrypoints/      CLI entry points
  platforms/        CUDA, NPU, and CPU abstractions
examples/           Runnable examples and model-specific usage guides
tests/              Unit, integration, and regression tests
docs/               English and Chinese documentation
tf-kernel/          Independently packaged optional CUDA kernels

Common Commands

pip install -e ".[dev]"
pre-commit run --all-files
pytest tests/
bash scripts/run_ci_tests.sh
telefuser serve /path/to/pipeline.py --port 8000
telefuser stream-serve /path/to/pipeline.py --port 8088

Read the full file on GitHub · 180 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 · 180 lines · 2,169 tokens per session scan A 3001dfb38743

Subscribe to this mod's changes

TeleFuser CLAUDE.md is an instructions file published in the GitHub repository Tele-AI/TeleFuser (24 stars, last pushed yesterday), licensed Apache-2.0. It adds 2,169 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.