Hacker News: zh2408

Show HN: 100-Line LLM Framework

zh2408 — Wed, 22 Oct 2025 22:33:28 +0000

Article URL: https://github.com/The-Pocket/PocketFlow

Comments URL: https://news.ycombinator.com/item?id=45676019

Points: 4

# Comments: 0

New comment by zh2408 in "Your LLM Framework Only Needs 100 Lines"

zh2408 — Sun, 13 Jul 2025 01:30:53 +0000

The 100 lines: https://github.com/The-Pocket/PocketFlow/blob/main/pocketflo... Video: https://www.youtube.com/watch?v=0Zr3NwcvpA0

Your LLM Framework Only Needs 100 Lines

zh2408 — Sun, 13 Jul 2025 01:30:53 +0000

Article URL: https://the-pocket.github.io/PocketFlow-Tutorial-Video-Generator/pocketflow

Comments URL: https://news.ycombinator.com/item?id=44546729

Points: 3

# Comments: 1

LLM Agents and Context: A Warrior's Guide to Navigating the Dungeon

zh2408 — Sun, 06 Jul 2025 19:32:57 +0000

Article URL: https://pocketflow.substack.com/p/llm-agents-and-context-a-warriors

Comments URL: https://news.ycombinator.com/item?id=44483341

Points: 4

# Comments: 0

LLM Agent's Arsenal: A Beginner's Guide to the Action Space

zh2408 — Wed, 02 Jul 2025 21:32:14 +0000

Article URL: https://pocketflow.substack.com/p/llm-agents-arsenal-a-beginners-guide

Comments URL: https://news.ycombinator.com/item?id=44449073

Points: 1

# Comments: 0

Show HN: I built an AI chatbot that learns from your website to answer questions

zh2408 — Sun, 29 Jun 2025 19:16:01 +0000

Article URL: https://github.com/The-Pocket/PocketFlow-Tutorial-Website-Chatbot

Comments URL: https://news.ycombinator.com/item?id=44415540

Points: 2

# Comments: 0

Structured Output for Beginners and 3 Prompting Tips

zh2408 — Fri, 20 Jun 2025 00:30:46 +0000

Article URL: https://pocketflow.substack.com/p/structured-output-for-beginners-3

Comments URL: https://news.ycombinator.com/item?id=44323779

Points: 2

# Comments: 0

Build AI Code Generator from Scratch

zh2408 — Sat, 24 May 2025 22:17:04 +0000

Article URL: https://pocketflow.substack.com/p/build-your-own-ai-code-generator

Comments URL: https://news.ycombinator.com/item?id=44084124

Points: 1

# Comments: 1

Build Your Own Voice Chatbot from Scratch

zh2408 — Thu, 15 May 2025 00:46:40 +0000

Article URL: https://pocketflow.substack.com/p/build-your-own-voice-chatbot-from

Comments URL: https://news.ycombinator.com/item?id=43990705

Points: 1

# Comments: 0

Show HN: LLM Agents Play Among Us-Like Game

zh2408 — Sun, 11 May 2025 15:06:08 +0000

Article URL: https://github.com/The-Pocket/PocketFlow-Tutorial-Danganronpa-Simulator

Comments URL: https://news.ycombinator.com/item?id=43954318

Points: 2

# Comments: 0

New comment by zh2408 in "Show HN: I built an AI that turns GitHub codebases into easy tutorials"

zh2408 — Wed, 23 Apr 2025 01:53:54 +0000

Could you try to use gemini 2.5 pro? It's free every day for first 25 requests, and can handle 1M input tokens

New comment by zh2408 in "Show HN: I built an AI that turns GitHub codebases into easy tutorials"

zh2408 — Sun, 20 Apr 2025 17:01:46 +0000

You can provide GitHub token

New comment by zh2408 in "Show HN: I built an AI that turns GitHub codebases into easy tutorials"

zh2408 — Sun, 20 Apr 2025 14:32:03 +0000

This code creates an HTTP session, sets a cookie within that session, makes another request that automatically includes the cookie, and then prints the response showing the cookies that were sent.

I may miss the error, but could you elaborate where it is?

New comment by zh2408 in "Show HN: I built an AI that turns GitHub codebases into easy tutorials"

zh2408 — Sun, 20 Apr 2025 12:57:55 +0000

Yeah, RAG is not the best option here. Check out the design doc: https://github.com/The-Pocket/Tutorial-Codebase-Knowledge/bl... I also have a YouTube Dev Tutorial. The link is on the repo.

New comment by zh2408 in "Show HN: I built an AI that turns GitHub codebases into easy tutorials"

zh2408 — Sun, 20 Apr 2025 12:51:17 +0000

By default we use both based on regex:

DEFAULT_INCLUDE_PATTERNS = { ".py", ".js", ".jsx", ".ts", ".tsx", ".go", ".java", ".pyi", ".pyx", ".c", ".cc", ".cpp", ".h", ".md", ".rst", "Dockerfile", "Makefile", ".yaml", ".yml", } DEFAULT_EXCLUDE_PATTERNS = { "test", "tests/", "docs/", "examples/", "v1/", "dist/", "build/", "experimental/", "deprecated/", "legacy/", ".git/", ".github/", ".next/", ".vscode/", "obj/", "bin/", "node_modules/", ".log" }

New comment by zh2408 in "Show HN: I built an AI that turns GitHub codebases into easy tutorials"

zh2408 — Sun, 20 Apr 2025 12:39:47 +0000

Thank you! And correct, I didn't modify the outputs. For small changes, you can just feed the commit history and ask an LLM to modify the docs. If there are lots of architecture-level changes, it would be easier to just feed the old docs and rewrite - it usually takes <10 minutes.

New comment by zh2408 in "Show HN: I built an AI that turns GitHub codebases into easy tutorials"

zh2408 — Sun, 20 Apr 2025 00:37:47 +0000

Thanks—would really appreciate your PR!

New comment by zh2408 in "Show HN: I built an AI that turns GitHub codebases into easy tutorials"

zh2408 — Sun, 20 Apr 2025 00:01:26 +0000

Their site seems to be down. I can't find their results.

New comment by zh2408 in "Show HN: I built an AI that turns GitHub codebases into easy tutorials"

zh2408 — Sat, 19 Apr 2025 23:49:06 +0000

Haha. The project is fully open-sourced, so you can tune the prompt for the tone/style you prefer: https://github.com/The-Pocket/Tutorial-Codebase-Knowledge/bl...

New comment by zh2408 in "Show HN: I built an AI that turns GitHub codebases into easy tutorials"

zh2408 — Sat, 19 Apr 2025 23:02:54 +0000

The Linux repository has ~50M tokens, which goes beyond the 1M token limit for Gemini 2.5 Pro. I think there are two paths forward: (1) decompose the repository into smaller parts (e.g., kernel, shell, file system, etc.), or (2) wait for larger-context models with a 50M+ input limit.