Borrowing it
Nothing to install: this file belongs to Tele-AI/TeleFuser. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Tele-AI/TeleFuser/main/.claude/skills/add-new-pipeline/SKILL.mdgit clone --depth 1 https://github.com/Tele-AI/TeleFuserWrote 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.
[](https://agentmods.dev/skills/tele-ai/telefuser/add-new-pipeline)<a href="https://agentmods.dev/skills/tele-ai/telefuser/add-new-pipeline"><img src="https://agentmods.dev/badge/skills/tele-ai/telefuser/add-new-pipeline/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.
<a href="https://agentmods.dev/skills/tele-ai/telefuser/add-new-pipeline"><img src="https://agentmods.dev/badge/skills/tele-ai/telefuser/add-new-pipeline.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00063 | $0.01066 |
| Opus 5 | $0.00032 | $0.00533 |
| Sonnet 5 | $0.00013 | $0.00213 |
| Haiku 4.5 | $0.00006 | $0.00107 |
Grade A, and why
add-new-pipeline 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 11d 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.
How it starts
The opening of the file, as written. The whole thing — 74 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Add a New Pipeline
Treat interface compatibility and upstream parity as hard requirements. Use the repository as the API source of truth; do not copy static API templates from this skill.
Establish the contract before editing
- Read the upstream entry point, model definitions, checkpoint configuration, preprocessing, scheduler loop, and output handling.
- Select the closest current TeleFuser pipeline, public example, and tests as structural baselines. Prefer a recently maintained implementation with the same task and loading pattern.
- Read only the relevant canonical documentation:
docs/en/adding_new_example.mddocs/en/adding_new_model.mddocs/en/adding_new_stage.mddocs/en/model_loading.mddocs/en/configuration.mddocs/en/service.mdwhen service support is in scope
- Inventory the interfaces the integration will reuse:
BasePipeline,BaseStage,ModuleManager, configuration dataclasses, example functions, contract/schema types, and service entry points. - Inventory required model-specific classes and configuration fields and map each one to upstream behavior and the selected baseline.
- List every proposed framework-level or cross-pipeline interface, general-purpose configuration field, environment variable, loader, registry, or service schema deviation. The expected list is empty. If one is genuinely necessary, explain why the existing extension points cannot represent it and obtain user approval before adding it.
Preserve upstream behavior
- Establish a minimal faithful path before splitting stages or optimizing operators.
- Preserve computation order, tensor shapes, conditioning paths, scheduler semantics, parameter meaning, default behavior, and output format.
- Read checkpoint metadata or configuration; never guess architecture dimensions or checkpoint-specific defaults.
- Keep preprocessing and postprocessing parity tests separate from end-to-end visual inspection.
- Do not combine integration with sparse attention, quantization, caching, refactoring, or performance tuning unless the user explicitly includes that work.
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.
- 11d ago First seen · 74 lines · 63 tokens per session scan A b73acbfd50e9
add-new-pipeline is a skill published in the GitHub repository Tele-AI/TeleFuser (26 stars, last pushed today), licensed Apache-2.0. It adds 63 tokens to every session and 1,066 once invoked, about $0.0003 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 skills, from other repositories
ml-training-recipes
Battle-tested PyTorch training recipes for all domains — LLMs, vision, diffusion, medical imaging, protein/drug discovery, spatial omics, genomics. Covers training loops, optimizer selection (AdamW, Muon), LR scheduling, mixed precision, debugging, and systematic experimentation. Use when training or fine-tuning…
ml-training-recipes
Battle-tested PyTorch training recipes for all domains — LLMs, vision, diffusion, medical imaging, protein/drug discovery, spatial omics, genomics. Covers training loops, optimizer selection (AdamW, Muon), LR scheduling, mixed precision, debugging, and systematic experimentation. Use when training or fine-tuning…
ml-training-recipes
Battle-tested PyTorch training recipes for all domains — LLMs, vision, diffusion, medical imaging, protein/drug discovery, spatial omics, genomics. Covers training loops, optimizer selection (AdamW, Muon), LR scheduling, mixed precision, debugging, and systematic experimentation. Use when training or fine-tuning…
tensorboard
Visualize training metrics, debug models with histograms, compare experiments, visualize model graphs, and profile performance with TensorBoard - Google's ML visualization toolkit.
ray-train
Distributed training orchestration across clusters. Scales PyTorch/TensorFlow/HuggingFace from laptop to 1000s of nodes. Built-in hyperparameter tuning with Ray Tune, fault tolerance, elastic scaling. Use when training massive models across multiple machines or running distributed hyperparameter sweeps.
torchforge-rl-training
Provides guidance for PyTorch-native agentic RL using torchforge, Meta's library separating infra from algorithms. Use when you want clean RL abstractions, easy algorithm experimentation, or scalable training with Monarch and TorchTitan.