robzolkos

A tiny Pi package that adds a CLI flag for printing Pi's assembled system prompt.

16
0
100% credibility
Found Apr 29, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

pi-nocchio is a small add-on for the Pi AI coding agent that lets users print and view the full assembled system prompt before any AI processing begins.

How It Works

1
🕵️‍♀️ Get curious about Pi's instructions

While chatting with your AI helper Pi, you wonder what full set of rules and tools it really uses to think and respond.

2
📥 Add the prompt peeker

You easily add this simple helper tool to your Pi setup so it can show you the hidden instructions anytime.

3
🔍 Ask Pi to reveal its prompt

You give Pi a special request to display the complete instructions it assembles before starting any work.

4
📄 See the full instructions

Pi instantly prints out the exact words, tools, date, and setup it uses, without doing any actual thinking or calling for help.

5
💾 Save the details

You copy or save the printed instructions to a file to review, share, or use later as needed.

Understand Pi perfectly

Now you clearly see everything Pi knows and uses, making it easier to guide or improve your AI conversations.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is pi-nocchio?

pi-nocchio is a tiny TypeScript package for the Pi coding agent CLI that adds a `--dump-system-prompt` flag. Run it with `pi --dump-system-prompt` to print Pi's fully assembled system prompt—including tools, context files, skills, custom prompts, date, and working directory—then exit before any model call. It solves the black-box frustration of debugging what Pi actually sends to the LLM, capturing turn-accurate output even from hooks, with a synthetic "dump" turn if no initial prompt is given.

Why is it gaining traction?

In a sea of github tiny projects like tiny c compiler or tiny shakespeare, this tiny package stands out for Pi users by delivering negligible perf overhead—just cheap flag checks—and full hook integration without model hits. Devs hook on its redirect-friendly stdout dump (`pi --dump-system-prompt > prompt.txt`) and GitHub install (`pi -e git:github.com/robzolkos/pi-nocchio`), making prompt inspection dead simple versus manual logging hacks. The "tiny package" vibe means zero bloat, pure utility.

Who should use this?

Pi extension authors verifying assembled prompts after their `before_agent_start` hooks. Devs tweaking custom system prompts, appends, or tools who need to see the real output. Teams auditing Pi sessions for context leaks or debugging why agents misbehave on specific turns.

Verdict

Grab it if you're deep in Pi workflows—solid docs and MIT license make it a no-risk add, despite 16 stars and 1.0% credibility score signaling early maturity. Skip for casual use; it's niche but nails its job.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.