What is tilelang-ascend AGENTS.md?

A project instruction file for TileLang-Ascend, a toolkit for writing AI computation kernels for Huawei Ascend chips with Python. It lists the project’s development skills, debugging tools, performance guidance, and working rules.

Use it as a starting reference when developing operators, compiler passes, APIs, tests, debugging setups, or performance improvements.

It gives coding agents the shared project context and directs them to the right instructions for designing, testing, debugging, and optimizing code.

How tile-ai/tilelang-ascend configures Codex and OpenCode

Reuse

Borrowing it

Nothing to install: this file belongs to tile-ai/tilelang-ascend. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/tile-ai/tilelang-ascend/ascendc_pto/AGENTS.md

Clone the repo

git clone --depth 1 https://github.com/tile-ai/tilelang-ascend

Made for: Codex, OpenCode.

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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/tile-ai/tilelang-ascend/agents-md/github.svg)](https://agentmods.dev/instructions/tile-ai/tilelang-ascend/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/tile-ai/tilelang-ascend/agents-md"><img src="https://agentmods.dev/badge/instructions/tile-ai/tilelang-ascend/agents-md/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.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/tile-ai/tilelang-ascend/agents-md"><img src="https://agentmods.dev/badge/instructions/tile-ai/tilelang-ascend/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 2,230 This file is loaded in full into every session.

When invoked 2,230 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.02230	$0.02230
Opus 5.5	$0.00892	$0.00892
Sonnet 5.5	$0.00446	$0.00446
Haiku 4.5	$0.00223	$0.00223

Measured today against content hash f14992a50876, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

tilelang-ascend AGENTS.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 today.

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.md · 130 lines

How it starts

The opening of the file, as written. The whole thing — 130 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

项目概述

本项目是 TileLang-Ascend 算子开发项目，基于 TVM 编译器基础设施，提供 Python DSL 用于开发华为昇腾 NPU 上的高性能 AI 计算 kernel。

核心功能

使用 Python DSL + @tilelang.jit 编写昇腾 NPU 自定义 kernel
支持 Developer 模式（自动化）和 Expert 模式（手动控制）两种编程范式
提供完整的编译、测试、调试及性能调优工作流

Skills 索引

算子开发与编排

tilelang-op-design：生成算子设计方案（design.md），含三维 Kernel / threads / 动态边界 / L0C 容量 / GEMM 非整除等技术约束检测
tilelang-op-develop：基于 design.md 生成算子实现代码、内嵌 golden 与测试用例
tilelang-ascend-tile-api：新增或封装 T.tile.xxx 小 API 时端到端打通前端、lowering / codegen、helper、测试与文档
tilelang-programming-model-guide：Developer / Expert / 混合模式选择、pass_configs 配置与转换指南
tilelang-api-best-practices：TileLang API 速查与最佳实践（Kernel 定义、内存分配、计算原语、调度同步）

Pass 分析与设计

tilelang-pass-analyzer：Pass 功能分析、对比、分类查询
tilelang-pass-workflow-analyzer：Pass 工作流分析、执行顺序、依赖关系、新 Pass 定位
tilelang-pass-design：Pass 设计方案与实现模式

调试与错误处理

tilelang-debug-helper：为算子添加 GDB 调试代码，配置 CMakeLists.txt 与 VSCode 联合调试
tilelang-error-fixer：编译 / 运行时错误诊断与修复

性能调优

tilelang-performance-best-practices：性能知识库，提供优化适用条件、实现约束、算子族参考与反模式检查

环境与工具

npu-arch：Ascend 架构与硬件能力查询；性能调优时结合已确认设备证据读取目标设备参数，并与编译器可用容量分别核验
tilelang-env-check：环境检查与配置验证（CANN、torch_npu、子模块、编译产物、环境变量）
tilelang-submodule-pull：自动拉取代码和子模块
tilelang-github-operations：GitHub PR 创建与操作

Skill 管理

skill-creator：创建新 skill
skill-journal：算子开发反馈记录 schema
tilelang-skill-review：聚合 skill-journal 反馈，按命令式 apply / reject 落到对应 SKILL.md
tilelang-review：通用 skill 质量评审

算子开发编排体系（OpenCode 多代理）

由 @tilelang-op-orchestrator 作为 Primary 驱动 3 阶段状态机，调度 3 个 Subagent：

@tilelang-op-analyst (Stage 1)：调用 tilelang-op-design 完成需求理解与设计
@tilelang-op-developer (Stage 2)：调用 tilelang-op-develop 完成代码实现、测试、精度调试（一站式，attempt 上限 5 次）
@tilelang-op-perf-tuner (Stage 3，可选)：负责数据采集、瓶颈分析、优化实施与复验，查阅 tilelang-performance-best-practices 核对实现约束

Read the full file on GitHub · 130 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.

today Changed · +1 lines · +66 tokens per session f14992a50876
1mo ago Changed · +23 tokens per session f329cf1fb0f3
2mo ago First seen · 129 lines · 2,141 tokens per session scan A c2f376b32df8

Subscribe to this mod's changes

tilelang-ascend AGENTS.md is an instructions file published in the GitHub repository tile-ai/tilelang-ascend (427 stars, last pushed yesterday), licensed MIT. It adds 2,230 tokens to every session, about $0.0089 per session on Opus 5.5. 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.

Data and AI Hardware and IoT

tilelang-ascend AGENTS.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
tilelang-ascend AGENTS.mdtile-ai/tilelang-ascend	Instructions file	427	yesterday	MIT	A	2,230	Codex, OpenCode
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	today	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	yesterday	MIT	A	4,680	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	5,516	GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	274k	2d ago	MIT	A	4,313	Codex, OpenCode

tilelang-ascend: Instructions file for Codex