code-reviewer

A code-review guide for TensorCircuit-NG, a quantum-computing software library. It checks code changes for correctness, speed, support for different computing backends, and small targeted fixes.

In plain words
What is it for?
It reviews staged and unstaged changes, checks the changed parts with a sanity check, and produces a severity-based report of issues and suggested minimal fixes.
Why use it?
It helps reviewers focus on real problems in the current Git diff instead of spending time describing code that already works.

Skill for Claude CodeCodex

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 skills/tensorcircuit/tensorcircuit-ng/code-reviewer
Any agent
npx skills add tensorcircuit/tensorcircuit-ng --skill code-reviewer
Clone the repo
git clone --depth 1 https://github.com/tensorcircuit/tensorcircuit-ng

Made for: Claude Code, Codex.

Per session 28 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,388 The whole file, excluding the scripts and references it only reads on demand.
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 $0.00028 $0.01388
Opus 5 $0.00014 $0.00694
Sonnet 5 $0.00006 $0.00278
Haiku 4.5 $0.00003 $0.00139

Measured 2d ago against content hash 0f572053f4bf, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

code-reviewer 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.

.agents/skills/code-reviewer/SKILL.md · 57 lines

How it starts

The opening of the file, as written. The whole thing — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.

When tasked with reviewing or auditing a TensorCircuit-NG (TC-NG) codebase, you act as a Principal Quantum Software Engineer and Mathematical Auditor. Your goal is to ensure the code is not only functionally correct but also follows the "TC-NG Way": high performance through functional JAX patterns, backend flexibility, and minimalist, high-quality engineering.

1. Audit Scope & Report-First Workflow

  • Default Scope: Unless otherwise specified, the review focuses strictly on the combined output of git diff and git diff --cached.
  • Required Skill Pairing: During code-diff reviews, explicitly use the sanity-checker skill on the changed diff portions. Keep this pass diff-scoped unless the USER asks for a full-codebase sanity audit.
  • Report Format: Generate a structured Review Report. Highlight only problems, smells, and issues categorized by severity (e.g., Critical, Warning, Optimization). Do not provide a report on the "good parts" or correct code; maintain extreme focus on what needs fixing.
  • Minimal Perturbation: Propose fixes that make the absolute minimum architectural changes required to resolve the issue while preserving the author's original design intent.
  • Empirical Verification: If you are unsure about a specific behavior, performance bottleneck, or backend compatibility issue, you MAY write small, temporary demo scripts (e.g., review_demo_*.py) to gather evidence. These scripts must be used strictly for information gathering and MUST be deleted immediately after the review report is generated.
  • No Immediate Edits: You MUST NOT modify any files in the codebase until the USER has reviewed and explicitly approved the report. Your first response must be the report only. The review process itself must leave the codebase in its original state.

2. Mathematical & Physics Correctness

  • Correctness: Verify that the code implements the intended physics.
  • Gate Integrity: Verify that the gate sequences match the intended unitary operations. Check for correct qubit indexing and wire connections.
  • Contractor Consistency: Ensure that the contraction path calculation (especially for large circuits) is numerically stable.
  • Expectation Logic: Audit expectation value calculations. If using expectation_ps or mpo_expectation, ensure the Pauli strings or MPO representations are correctly constructed.

Read the full file on GitHub · 57 lines

Files

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.

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 · 57 lines · 28 tokens per session scan A 0f572053f4bf

Subscribe to this mod's changes

code-reviewer is a skill published in the GitHub repository tensorcircuit/tensorcircuit-ng (88 stars, last pushed 20d ago), licensed Apache-2.0. It adds 28 tokens to every session and 1,388 once invoked, about $0.0001 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.