AgentStack
i-stack avatar

i-stack

12 listings · 0 installs

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

↗ github.com/i-stack
12 results
Self-run
MCP

Ai Coding Kit

One-kit AI coding workflow — Agent Skills, MCP config sync, iOS engineering rules & RAG Gateway for Cursor, Codex, Claude Code, CodeBuddy, Gemini, Continue, Cline & Xcode

0
5
Free
Self-run
SKILL

Ios Engineer

iOS / Swift / SwiftUI / UIKit / Xcode / CocoaPods / SPM engineering - architecture, concurrency, networking, performance, crash debugging, code review, refactoring, migration, testing. Covers design, implementation, and production risk control.

0
1
Free
Self-run
SKILL

References

A Claude skill from i-stack/ai-coding-kit.

0
0
Free
Self-run
SKILL

Cognitive Expansion

>-

0
1
Free
Self-run
SKILL

Ios Engineer

iOS / Swift / SwiftUI / UIKit / Xcode / CocoaPods / SPM engineering - architecture, concurrency, networking, performance, crash debugging, code review, refactoring, migration, testing. Covers design, implementation, and production risk control.

0
0
Free
Self-run
SKILL

Problem Analysis

问题前置分析——逻辑检验、第一性原理拆解、充分理解后再回复(PA-001/002/003)。适用所有含判断或方案讨论的任务。

0
0
Free
Self-run
SKILL

Epistemic Integrity

全局真值接地纪律——不把未验证内容说成已知、自信≠正确、逼出可验证物、验证方法论与求真方法边界(GR-011/012/013)。适用所有含事实性断言或解惑型回答的任务,不限平台。

0
1
Free
Self-run
SKILL

Logical Reasoning

全局论证纪律——可追溯逻辑链、层级分明、因果克制、逻辑链输出块(GR-010)。适用所有工程任务,不限平台。

0
1
Free
Self-run
SKILL

Engineering Discipline

全局工程纪律——安全合规防御、前置确认、单根因、四段式、最小修复、预算拦截、防Diff噪声、残留风险声明(GR-001...008)。适用所有工程任务,不限平台。

0
1
Free
Self-run
SKILL

Plan Grill

需求对齐/盘问锁定计划。收到非平凡构建、修改或方案请求时,先评估是否存在无法从代码或上下文确定、且会实质改变结果的阻塞性决策;有则自动进入逐问盘问,无则直接回复或执行。显式 grill/锁定计划触发语始终强制进入。确认前不执行,产出 PLAN.md 供 cross-model-review 接力。基于 Matt Pocock 的 grilling(MIT)。

0
2
Free
Self-run
SKILL

Cross Model Review

跨模型对抗审查已锁定的 PLAN.md——自动发现可用 CLI(codex/gemini/claude),推荐两个不同 provider 的组合并让用户选择,reviewer 只读运行输出 VERDICT:APPROVED|REVISE,主 agent 仲裁并把理由写进 PLAN-REVIEW-LOG.md,MAX_ROUNDS 不收敛输出 deadlock。基于 chaseai-yt/grill-me-codex(MIT)的 Act 2 思路。

0
5
Free
Self-run
SKILL

Auto Code Review

用户显式触发的跨模型代码审查工作流。仅当用户明确说 `/auto-review`、`使用 auto-code-review`、`启动跨模型代码审查`,或明确要求“审查并修复”时使用;普通代码生成、修改完成或含糊的“看看代码”不自动触发。默认只读审查,只有用户明确要求 `--fix` 或“审查并修复”才允许主 agent 修改代码。

0
2
Free
You've reached the end · 12 loaded