AgentStack
Nothern131 avatar

Nothern131

6 listings · 0 installs

Open-source publisher. Listings imported from github.com/Nothern131 — credited to the original author with their license.

↗ github.com/Nothern131
6 results
Self-run
SKILL

Codebase Indexer

Generates a codebase map for domestic AI models to compensate for lack of native codebase indexing. Scans project structure, identifies key modules, entry points, and dependency relationships. Use when starting a new project, when the user says 'explore/了解/看看/结构', or when the model needs to understand a large codebase quickly. Outputs a structured PROJECT_CONTEXT.md with directory tree, module de…

0
3
Free
Self-run
SKILL

Test Automation

Auto-detects test framework, runs tests, parses results, and feeds back to the Closed Loop for autonomous quality verification. Compensates for domestic models' lack of native test execution. Use after code generation, after Closed Loop fixes, or when the user says 'test/验证/跑测试'. Works with pytest, unittest, Jest, Go test, and GDScript GUT.

0
0
Free
Self-run
SKILL

Clarifying Questions

Helps AI precisely understand user requirements by asking targeted clarifying questions before starting work. Use this skill whenever a user's request is vague, ambiguous, or lacks critical details — such as missing tech stack, unclear scope, unspecified format, or undefined constraints. Also use proactively when the user says things like 'build me something', 'make it better', 'fix it', or any r…

0
2
Free
Self-run
SKILL

Template Skill

Replace with description of the skill and when the AI assistant should use it.

0
5
Free
Self-run
SKILL

Model Upgrade

Engineering framework v3 that enables domestic AI models (GLM, DeepSeek, Qwen, etc.) to achieve GPT/Codex-level output quality through a three-layer architecture: Short Constitution (10 iron rules), Capability Enhancement (reasoning, diagnosis, consistency), and Execution Templates (fixed formats). Designed for the TRAE CN ecosystem. Core formula: Quality = Model_Raw × Engineering_Multiplier.

0
2
Free
Self-run
SKILL

Codeguard

Automated code quality scanner that detects 80 common defects across 8 dimensions (error handling, security, database protection, diagnostics, resilience, memory safety, concurrency, resource management). Covers Python, JavaScript/TypeScript, C/C++, GDScript, and Go. Use this skill whenever the user asks for code review, code quality check, code audit, security scan, or wants to ensure their code…

0
2
Free
You've reached the end · 6 loaded