blockrun
Use when user needs capabilities Claude lacks (image generation, real-time X/Twitter data) or explicitly requests external models ("blockrun", "use grok", "use gpt", "dall-e", "deepseek")
- License 许可
- MIT
- 最近更新
- 2026-10-13
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- blockrun
- description
- Use when user needs capabilities Claude lacks (image generation, real-time X/Twitter data) or explicitly requests external models ("blockrun", "use grok", "use gpt", "dall-e", "deepseek")
- allowed-tools
- Read, Bash(python:*), Bash(python3:*), Bash(pip:*), Bash(source:*)
03 SKILL.md 正文
BlockRun
BlockRun works with Claude Code and Google Antigravity.
You can't generate images. You have no real-time X/Twitter access. BlockRun gives you these capabilities.
Philosophy
You have a wallet. When you need a capability you lack, you pay for it autonomously.
| You Need | You Call | Pricing |
|---|---|---|
| Image generation | DALL-E | $0.04/image |
| Real-time X data | Grok + Live Search | $0.025/source |
| Second opinion | GPT-5.2 | $1.75/M input, $14/M output |
| Cheaper processing | DeepSeek | $0.14/M input, $0.28/M output |
How it works: BlockRun uses x402 micropayments to route your requests to OpenAI, xAI, Google, and other providers. No API keys needed - your wallet pays per token.
Budget Control (Optional)
If the user specifies a budget (e.g., "use at most $1"), track spending and stop when budget is reached:
from blockrun_llm import setup_agent_wallet
client = setup_agent_wallet()
budget = 1.0 # User's budget
# Before each call, check if within budget
spending = client.get_spending()
if spending['total_usd'] >= budget:
print(f"Budget reached: ${spending['total_usd']:.4f} spent")
# Stop making calls
else:
response = client.chat("openai/gpt-5.2", "Hello!")
# At the end, report spending
spending = client.get_spending()
print(f"💰 Total spent: ${spending['total_usd']:.4f} across {spending['calls']} calls")
When to Use
| Trigger | Your Action |
|---|---|
| User explicitly requests ("blockrun second opinion with GPT on...", "use grok to check...", "generate image with dall-e") | Execute via BlockRun |
| User needs something you can't do (images, live X data) | Suggest BlockRun, wait for confirmation |
| You can handle the task fine | Do it yourself, don't mention BlockRun |
Example User Prompts
Users will say things like:
| User Says | What You Do |
|---|---|
| "blockrun generate an image of a sunset" | Call DALL-E via ImageClient |
| "use grok to check what's trending on X" | Call Grok with search=True |
| "blockrun GPT review this code" | Call GPT-5.2 via LLMClient |
| "what's the latest news about AI agents?" | Suggest Grok (you lack real-time data) |
| "generate a logo for my startup" | Suggest DALL-E (you can't generate images) |
| "blockrun check my balance" | Show wallet balance via get_balance() |
| "blockrun deepseek summarize this file" | Call DeepSeek for cost savings |
Wallet & Balance
Use setup_agent_wallet() to auto-create a wallet and get a client. This shows the QR code and welcome message on first use.
Initialize client (always start with this):
from blockrun_llm import setup_agent_wallet
client = setup_agent_wallet() # Auto-creates wallet, shows QR if new
Check balance (when user asks "show balance", "check wallet", etc.):
balance = client.get_balance() # On-chain USDC balance
print(f"Balance: ${balance:.2f} USDC")
print(f"Wallet: {client.get_wallet_address()}")
Show QR code for funding:
from blockrun_llm import generate_wallet_qr_ascii, get_wallet_address
# ASCII QR for terminal display
print(generate_wallet_qr_ascii(get_wallet_address()))
SDK Usage
Prerequisite: Install the SDK with pip install blockrun-llm
Basic Chat
from blockrun_llm import setup_agent_wallet
client = setup_agent_wallet() # Auto-creates wallet if needed
response = client.chat("openai/gpt-5.2", "What is 2+2?")
print(response)
# Check spending
spending = client.get_spending()
print(f"Spent ${spending['total_usd']:.4f}")
Real-time X/Twitter Search (xAI Live Search)
IMPORTANT: For real-time X/Twitter data, you MUST enable Live Search with search=True or search_parameters.
from blockrun_llm import setup_agent_wallet
client = setup_agent_wallet()
# Simple: Enable live search with search=True
response = client.chat(
"xai/grok-3",
"What are the latest posts from @blockrunai on X?",
search=True # Enables real-time X/Twitter search
)
print(response)
Advanced X Search with Filters
from blockrun_llm import setup_agent_wallet
client = setup_agent_wallet()
response = client.chat(
"xai/grok-3",
"Analyze @blockrunai's recent content and engagement",
search_parameters={
"mode": "on",
"sources": [
{
"type": "x",
"included_x_handles": ["blockrunai"],
"post_favorite_count": 5
}
],
"max_search_results": 20,
"return_citations": True
}
)
print(response)
Image Generation
from blockrun_llm import ImageClient
client = ImageClient()
result = client.generate("A cute cat wearing a space helmet")
print(result.data[0].url)
xAI Live Search Reference
Live Search is xAI's real-time data API. Cost: $0.025 per source (default 10 sources = ~$0.26).
To reduce costs, set max_search_results to a lower value:
# Only use 5 sources (~$0.13)
response = client.chat("xai/grok-3", "What's trending?",
search_parameters={"mode": "on", "max_search_results": 5})
Search Parameters
| Parameter | Type | Default | Description |
|---|---|---|---|
mode |
string | "auto" | "off", "auto", or "on" |
sources |
array | web,news,x | Data sources to query |
return_citations |
bool | true | Include source URLs |
from_date |
string | - | Start date (YYYY-MM-DD) |
to_date |
string | - | End date (YYYY-MM-DD) |
max_search_results |
int | 10 | Max sources to return (customize to control cost) |
Source Types
X/Twitter Source:
{
"type": "x",
"included_x_handles": ["handle1", "handle2"], # Max 10
"excluded_x_handles": ["spam_account"], # Max 10
"post_favorite_count": 100, # Min likes threshold
"post_view_count": 1000 # Min views threshold
}
Web Source:
{
"type": "web",
"country": "US", # ISO alpha-2 code
"allowed_websites": ["example.com"], # Max 5
"safe_search": True
}
News Source:
{
"type": "news",
"country": "US",
"excluded_websites": ["tabloid.com"] # Max 5
}
Available Models
| Model | Best For | Pricing |
|---|---|---|
openai/gpt-5.2 |
Second opinions, code review, general | $1.75/M in, $14/M out |
openai/gpt-5-mini |
Cost-optimized reasoning | $0.30/M in, $1.20/M out |
openai/o4-mini |
Latest efficient reasoning | $1.10/M in, $4.40/M out |
openai/o3 |
Advanced reasoning, complex problems | $10/M in, $40/M out |
xai/grok-3 |
Real-time X/Twitter data | $3/M + $0.025/source |
deepseek/deepseek-chat |
Simple tasks, bulk processing | $0.14/M in, $0.28/M out |
google/gemini-2.5-flash |
Very long documents, fast | $0.15/M in, $0.60/M out |
openai/dall-e-3 |
Photorealistic images | $0.04/image |
google/nano-banana |
Fast, artistic images | $0.01/image |
M = million tokens. Actual cost depends on your prompt and response length.
Cost Reference
All LLM costs are per million tokens (M = 1,000,000 tokens).
| Model | Input | Output |
|---|---|---|
| GPT-5.2 | $1.75/M | $14.00/M |
| GPT-5-mini | $0.30/M | $1.20/M |
| Grok-3 (no search) | $3.00/M | $15.00/M |
| DeepSeek | $0.14/M | $0.28/M |
| Fixed Cost Actions | |
|---|---|
| Grok Live Search | $0.025/source (default 10 = $0.25) |
| DALL-E image | $0.04/image |
| Nano Banana image | $0.01/image |
Typical costs: A 500-word prompt (~750 tokens) to GPT-5.2 costs ~$0.001 input. A 1000-word response (~1500 tokens) costs ~$0.02 output.
Setup & Funding
Wallet location: $HOME/.blockrun/.session (e.g., /Users/username/.blockrun/.session)
First-time setup:
- Wallet auto-creates when
setup_agent_wallet()is called - Check wallet and balance:
from blockrun_llm import setup_agent_wallet
client = setup_agent_wallet()
print(f"Wallet: {client.get_wallet_address()}")
print(f"Balance: ${client.get_balance():.2f} USDC")
- Fund wallet with $1-5 USDC on Base network
Show QR code for funding (ASCII for terminal):
from blockrun_llm import generate_wallet_qr_ascii, get_wallet_address
print(generate_wallet_qr_ascii(get_wallet_address()))
Troubleshooting
"Grok says it has no real-time access"
→ You forgot to enable Live Search. Add search=True:
response = client.chat("xai/grok-3", "What's trending?", search=True)
Module not found
→ Install the SDK: pip install blockrun-llm
Updates
pip install --upgrade blockrun-llm
04 同仓库更多 Skills
zapier-make-patterns
No-code automation democratizes workflow building. Zapier and Make (formerly Integromat) let non-developers automate business processes without writing code. But no-code doesn't mean no-complexity - these platforms have their own patterns, pitfalls, and breaking points. This skill covers when to use which platform, how to build reliable automations, and when to graduate to code-based solutions. Key insight: Zapier optimizes for simplicity and integrations (7000+ apps), Make optimizes for power
yeet
Use only when the user explicitly asks to stage, commit, push, and open a GitHub pull request in one flow using the GitHub CLI (`gh`).
workflow-automation
Workflow automation is the infrastructure that makes AI agents reliable. Without durable execution, a network hiccup during a 10-step payment flow means lost money and angry customers. With it, workflows resume exactly where they left off. This skill covers the platforms (n8n, Temporal, Inngest) and patterns (sequential, parallel, orchestrator-worker) that turn brittle scripts into production-grade automation. Key insight: The platforms make different tradeoffs. n8n optimizes for accessibility
trigger-dev
Trigger.dev expert for background jobs, AI workflows, and reliable async execution with excellent developer experience and TypeScript-first design. Use when: trigger.dev, trigger dev, background task, ai background job, long running task.
slack-automation
Automate Slack workspace operations including messaging, search, channel management, and reaction workflows through Composio's Slack toolkit.
rote
Compile a proven agent skill (a SKILL.md plus references) into a deterministic pipeline that runs without an LLM in the loop, then serve it back to Claude as an MCP tool. Use when: rote, compile this skill, turn this skill into a workflow, make this skill deterministic, make this skill cheaper or faster, harden this skill for production, run this skill as a background job.