Conversation
There was a problem hiding this comment.
💡 Codex Review
spoon-toolkit/requirements.txt
Lines 12 to 16 in f4647b7
Installing from requirements.txt now skips neo3>=1.0.0 (it was removed in this change), but the Neo modules still import neo3.api/neo3.core (spoon_toolkits/crypto/neo/neo_provider.py lines 22–25). Any environment built from this requirements file will raise ModuleNotFoundError: neo3 as soon as a Neo tool is imported, so the dependency needs to be re-listed.
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Summary
This PR introduces a new, reusable
WebScraperToolto thespoon-toolkitlibrary.Previously, examples like
x402_agent_demo.pyused ad-hoc, primitive HTTP getters (HttpProbeTool) that returned raw HTML. This new tool provides a robust, production-ready solution designed specifically for LLM agents, featuring HTML-to-Markdown conversion, content cleaning, and x402 protocol compatibility.Key Features
beautifulsoup4andmarkdownifyto convert raw HTML into clean Markdown, significantly reducing token usage and noise.<script>,<style>,<iframe>, and heuristic-based ad elements.402 Payment Requiredresponses by returning the payment headers (instead of raising exceptions), enabling agents to detect and pay for content seamlessly.httpxfor non-blocking performance.Changes
spoon_toolkits/web/web_scraper.py.spoon_toolkits/web/__init__.py.beautifulsoup4andmarkdownifytopyproject.toml/requirements.txt.HttpProbeToolinexamples/x402_agent_demo.pywith this new standard tool.