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.
npx agentmods add skills/tensorcircuit/tensorcircuit-ng/tc-rosettanpx skills add tensorcircuit/tensorcircuit-ng --skill tc-rosettagit clone --depth 1 https://github.com/tensorcircuit/tensorcircuit-ngWrote 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/tensorcircuit/tensorcircuit-ng/tc-rosetta)<a href="https://agentmods.dev/skills/tensorcircuit/tensorcircuit-ng/tc-rosetta"><img src="https://agentmods.dev/badge/skills/tensorcircuit/tensorcircuit-ng/tc-rosetta.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00056 | $0.00881 |
| Opus 5 | $0.00028 | $0.00441 |
| Sonnet 5 | $0.00011 | $0.00176 |
| Haiku 4.5 | $0.00006 | $0.00088 |
Grade A, and why
tc-rosetta 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 6d 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 — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.
When tasked with translating quantum computing code from frameworks like Qiskit, PennyLane, or Cirq into TensorCircuit-NG (TC-NG), you act as a Principal Quantum Software Migration Expert.
Your goal is NOT to do a naive 1:1 syntax or line-by-line translation. You must perform End-to-End Intent Understanding to refactor the code into the idiomatic, differentiable, and functional programming paradigm of TC-NG.
1. End-to-End Intent Extraction (Do NOT Translate Line-by-Line)
- Read the Entire Script: Absorb the global objective of the source code (e.g., VQE, QAOA, QFI).
- Extract the Math/Physics Core: Extract only the fundamental mathematical entities: the Ansatz architecture, the target Hamiltonian, and the loss function.
- Discard Legacy Paradigms: Explicitly abandon the original framework's constraints (e.g., Qiskit's parameter-binding
forloops).
2. Idiomatic Synthesis in TC-NG
- Start Fresh: Write the TC-NG script from scratch based on the extracted intent.
- JAX-Native Initialization:
import tensorcircuit as tc import jax import jax.numpy as jnp import time tc.set_backend("jax") - Functional Paradigm & Vectorization: Construct the circuit execution as a pure, differentiable Python function. Apply
tc.backend.vmapnatively for any batched operations. - Programming Paradigms: Avoid over-defensive programming; trust internal invariants where reasonable. Use
try...exceptsparingly and never use broad catch-all blocks likeexcept Exception:. Fail fast and expose problems early rather than masking them with silent failures or broad error handling.
3. Execution, Verification & Strict Benchmarking
- Run the Source Code: If the original script is executable, run it and strictly record its total execution time using
timeor a simple bashtime python script.py. - Run the TC-NG Code: Execute your newly synthesized TC-NG script. Ensure it outputs numerically equivalent results (e.g., matching loss curves or final energies).
- Record Performance: Measure the exact execution time of the TC-NG script, noting separately the JIT compilation time (if applicable/measurable) and the actual execution time.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 6d ago First seen · 48 lines · 56 tokens per session scan A d302c3c734a3
tc-rosetta is a skill published in the GitHub repository tensorcircuit/tensorcircuit-ng (89 stars, last pushed 4d ago), licensed Apache-2.0. It adds 56 tokens to every session and 881 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
physicsnemo-discover
Official NVIDIA-authored guidance for navigating PhysicsNeMo — pick the model, datapipe, or example for a SciML/AI4Science task (surrogates, forecasting, downscaling, physics-informed, inverse, generative). Points at existing files via live repo search; never writes code. Do NOT use for installation or environment…
cellxgene-census-query
Query CZ CELLxGENE Census (61M+ cells). Filter by cell type/tissue/disease, retrieve expression data, and integrate with scanpy/PyTorch for population-scale single-cell analysis. Use this skill when: (1) Querying single-cell expression data by cell type, tissue, or disease, (2) Exploring available single-cell datasets…
start-experiment
Start the autoresearch optimization loop for a specific model + lane. Resolves the hierarchical program.md (root → model → lane), asks the user for hardware (local TPU VM or GKE cluster of a specified TPU type + topology), discovers available clusters from .env/, checks occupancy with USERPREFIX-aware attribution…
stop-experiment
Stop the autoresearch optimization loop cleanly. Disarms the launch-armed process watcher, cancels pending wakeups, reaps orphan workloads on the project's cluster pool, files any missing experiment pages, runs lint, and appends a clean-shutdown marker to wiki/log.md. Invoke at the end of an autoresearch session.
edit-model-code
Coding-discipline guidelines for surgically modifying model code inside a per-experiment fork in the autoresearch loop. Adapts Andrej Karpathy's LLM-coding guidelines (think before coding, simplicity first, surgical changes, goal-driven execution) to this project's invariants — no semantics drift, one mechanism per…
author-kernel
Authoring discipline for kernel-family candidates (the author role of kernelexperiments/program.md step K4). Load BEFORE writing any kernel candidate — whether you are a solo agent running the whole K0–K9 loop (production: Gemini / Claude / Codex from a bare prompt) or a dispatched author process in the trial…