Built-in tools

> what dependencies does my application have

Reading file: src/snake/package.json, all lines (using tool: read)
 ✓ Successfully read 1417 bytes from src/snake/package.json
 - Completed in 0.86s

> Your application has:

Runtime Dependencies:
- typescript (^3.5.1) - TypeScript compiler
- gh-pages (^2.0.1) - GitHub Pages deployment

Dev Dependencies:
- sass (^1.20.3) - CSS preprocessing
- eslint (^5.16.0) + TypeScript plugins - Code linting
- jest (^29.5.0) + ts-jest - Testing framework

No external game libraries or frameworks.

{
  "toolsSettings": {
    "read": {
      "allowedPaths": ["~/projects", "./src/**"],
      "deniedPaths": ["d1/denied/path/", "d2/denied/path/**/file.txt"]
    }
  }
}

> find all TypeScript test files in my project

Finding files matching pattern: **/*.test.ts (using tool: glob)

 ✓ Found 8 files
 - Completed in 0.12s

> Here are the TypeScript test files:

src/utils/helpers.test.ts
src/utils/validators.test.ts
src/components/button.test.ts
src/components/header.test.ts
src/services/auth.test.ts
src/services/api.test.ts
src/hooks/useAuth.test.ts
src/hooks/useForm.test.ts

{
  "toolsSettings": {
    "glob": {
      "allowedPaths": ["~/projects", "./src/**"],
      "deniedPaths": ["/etc", "/var"],
      "allowReadOnly": true
    }
  }
}

> find all TODO comments in my project

Searching for pattern: TODO (using tool: grep)

 ✓ Found 12 matches across 5 files
 - Completed in 0.34s

> Here are the TODO comments I found:

src/utils/helpers.ts:23
  // TODO: Add error handling for edge cases

src/components/header.tsx:45
  {/* TODO: Implement mobile navigation */}

src/api/routes.ts:12
  // TODO: Add rate limiting
  // TODO: Implement caching

src/services/auth.ts:78
  // TODO: Add refresh token logic

...

{
  "toolsSettings": {
    "grep": {
      "allowedPaths": ["~/projects", "./src/**"],
      "deniedPaths": ["/etc", "/var"],
      "allowReadOnly": true
    }
  }
}

> collect best practices for writing typescript code and save it to the kiro steering folder as codingpractices.md

I'll create the following file: src/snake/.kiro/steering/codingpractices.md (using tool: write)
Purpose: Create TypeScript coding best practices document

+     1: # TypeScript Coding Best Practices
+     2: 
+     3: ## Type Safety
+     4: - Enable strict mode in `tsconfig.json` (`strict: true`)
+     5: - Avoid `any` type - use `unknown` for truly dynamic types
+     6: - Use explicit return types for functions
+     7: - Leverage type inference where it improves readability
+     8: - Use union types instead of enums when possible
+     9: - Prefer `interface` for object shapes, `type` for unions/intersections
+    10: 
+    11: ## Code Organization
+    12: - One class/interface per file (exceptions for small related types)
+    13: - Group related functionality into modules
+    14: - Use barrel exports (`index.ts`) for clean imports
+    15: - Keep functions small and focused (single responsibility)
+    16: - Separate business logic from presentation logic

{
  "toolsSettings": {
    "write": {
      "allowedPaths": ["~/projects/output.txt", "./src/**"],
      "deniedPaths": ["/d1/denied/path/", "/d2/denied/path/**/file.txt"]
    }
  }
}

{
  "toolsSettings": {
    "shell": {
      "allowedCommands": ["git status", "git fetch"],
      "deniedCommands": ["git commit .*", "git push .*"],
      "autoAllowReadonly": true
    }
  }
}

 list my s3 buckets

Running aws cli command (using tool: aws):

Service name: s3api
Operation name: list-buckets
Parameters: 
Region: us-east-1
Label: List all S3 buckets

{
  "toolsSettings": {
    "aws": {
      "allowedServices": ["s3", "lambda", "ec2"],
      "deniedServices": ["eks", "rds"],
      "autoAllowReadonly": true
    }
  }
}

# Example 1
> What is the latest on EC2 instances?

> I'll search for the latest information on EC2 instances.
Searching the web for: AWS EC2 instances latest 2025 (using tool: web_search)

 ✓ Found 10 search results
 - Completed in 2.12s

> Here's what I found:

Recent EC2 Instance Launches (2025):

New Instance Families:
- **M8a instances** - General-purpose instances with 5th Gen AMD EPYC (Turin) processors running at 4.5GHz. They deliver up to 30% higher performance and 19% better price-performance versus M7a instances.

- **C8gd, M8gd, R8gd instances** - Graviton-based instances with up to 11.4 TB of local NVMe SSD storage, now generally available.

- **X8g instances** - Powered by AWS Graviton4 processors, delivering up to 60% better performance than X2gd instances. Available in US East (Ohio).

- **R8i and R8i-flex instances** - Intel-based memory-optimized instances offering up to 15% better price-performance and 2.5x more memory bandwidth compared to previous Intel generations. They deliver 20% better performance than R7i 
instances.

- **I7i instances** - Storage-optimized instances with 5th Gen Intel Xeon processors, delivering up to 23% better compute performance and 10% better price-performance over I4i instances.

Expanded Availability:
- C7i-flex and M7i-flex now available in 12xlarge and 16xlarge sizes
- Various instances expanding to additional regions including Europe (Milan) and US West (N. California)

Key Trend: AWS continues focusing on performance improvements (20-30% gains), better price-performance (10-19% improvements), and expanding processor options across Intel, AMD, and Graviton architectures.

References:
[1] Document history for the Amazon EC2 Instance Types Guide - https://docs.aws.amazon.com/ec2/latest/instancetypes/doc-history.html
... 
[10] Amazon EC2 C7i-flex and C7i instances are now available in 2 additional regions - https://aws.amazon.com/about-aws/whats-new/2025/06/amazon-ec2-c7i-flex-c7i-instances-additional-...
 ▸ Time: 19s


# Example 2
> What is the latest on t3 medium instance pricing.

Searching the web for: AWS EC2 t3.medium instance pricing 2025 (using tool: web_search)

 ✓ Found 10 search results
 - Completed in 1.596s

Fetching content from: https://aws.amazon.com/ec2/pricing/on-demand/ (searching for: t3.medium price cost) [mode: selective] (using tool: web_fetch)

 ✓ Fetched 11395 bytes (selective) from URL
 - Completed in 0.718s

> The page doesn't show specific pricing in the excerpt. Let me get the actual pricing information:
Fetching content from: https://instances.vantage.sh/aws/ec2/t3.medium (searching for: t3.medium $0.04) [mode: selective] (using tool: web_fetch)

 ✓ Fetched 3971 bytes (selective) from URL
 - Completed in 0.780s

> Here's what I found:

AWS EC2 t3.medium Instance Pricing (2025):

...

# Example - using truncated
>  https://kiro.dev/blog/introducing-kiro-cli/ --> Can you get some of this page for me?

Fetching content from: https://kiro.dev/blog/introducing-kiro-cli/ [mode: truncated] (using tool: web_fetch)

 ✓ Fetched 8051 bytes (truncated content) from URL
 - Completed in 0.521s

> I've fetched the beginning portion of the page. It's a blog post announcing Kiro CLI. Here's what this section covers:

Title: "Bring Kiro agents to your terminal with Kiro CLI" (November 17, 2025)

Main Points:
- Kiro CLI brings AI agents into your terminal so you don't lose context when debugging production servers or working in SSH sessions
- Built on Q Developer CLI technology with additions like social login, Haiku 4.5, and the Auto agent
- Supports custom agents tailored to specific workflows (backend, frontend, DevOps)
- Your .kiro folder configs work in both IDE and CLI

Installation:
bash
curl -fsSL https://cli.kiro.dev/install | bash


Key Features covered in this section:
1. Custom agents with pre-approved tools, persistent context, and controlled access
2. Smart context management (Agent Resources, Session Context, Knowledge Bases)
3. Multiple authentication options (GitHub, Google, AWS Builder ID, IAM Identity Center)
4. Full IDE integration - MCP servers and steering rules transfer seamlessly
5. Interactive chat, multimodal input, and MCP support

The page continues beyond what was fetched. Would you like me to get a specific section or search for particular information?

# Example - using selective 

> https://kiro.dev/blog/introducing-kiro-cli/ --> Can you get installation information from this page for me.

Fetching content from: https://kiro.dev/blog/introducing-kiro-cli/ (searching for: installation install getting started) [mode: selective] (using tool: web_fetch)

 ✓ Fetched 7909 bytes (selective) from URL
 - Completed in 0.434s

> Here's the installation information from the page:

## Installation

Kiro CLI is available for macOS and Linux. Installation is straightforward:

bash
curl -fsSL https://cli.kiro.dev/install | bash


## First Steps

1. Authenticate and start chatting: Sign in with your credentials

2. Explore commands: Get help anytime

The page also mentions at the end:

bash
curl -fsSL https://cli.kiro.dev/install | bash


That's the single command you need to install Kiro CLI on macOS or Linux systems.

{
  "toolsSettings": {
    "web_fetch": {
      "trusted": [".*docs\\.aws\\.amazon\\.com.*", ".*github\\.com.*"],
      "blocked": [".*pastebin\\.com.*"]
    }
  }
}

> How do I save conversations?

Introspecting to get you the right information (using tool: introspect) - Completed in 0.68s

> You can save conversations using `/chat save <PATH>`:

- `/chat save ~/conversation.json` - Save to a specific path

Load saved conversations later with `/chat load <PATH>`.

kiro-cli settings set introspect.progressiveMode true

> What experimental features does Kiro CLI have?
> Can Kiro CLI read and write files?
> How do I configure MCP servers?
> What settings are available?

kiro-cli settings set introspect.tangentMode true

> Find the UserRepository class

Searching for symbols matching: "UserRepository" (using tool: code)

 ✓ Found 1 match
 - Completed in 0.45s

> Found:
  Class UserRepository at src/repositories/user.repository.ts:15:1

> Analyze all TypeScript files for potential bugs and create a report

Delegating task to background agent (using tool: delegate)

 ✓ Task delegated successfully
 - Agent ID: agent-abc123
 - Check status with /delegate status

> I've started analyzing your TypeScript files in the background. 
  Use /delegate status to check progress.

> Research the top 3 JavaScript frameworks and compare their performance

# Main agent spawns 3 subagents:
# - Subagent 1: Research React performance metrics
# - Subagent 2: Research Vue.js performance metrics
# - Subagent 3: Research Angular performance metrics

# Each subagent:
# - Conducts independent research
# - Gathers relevant data
# - Calls summary tool with findings

# Main agent receives all summaries and synthesizes comparison

{
  "toolsSettings": {
    "write": {
      "allowedPaths": ["~/projects"]
    },
    "@git/git_status": {
      "git_user": "$GIT_USER"
    }
  }
}

{
  "allowedTools": [
    "read",
    "knowledge",
    "@git/git_status"
  ]
}