# ZiChuanLan/pi-purge

> A pi extension whose /purge command swaps the session's system prompt for your own file or a bundled example in append, replace, or strict mode. Built for prompt-engineering experiments and regression-testing your instructions.

- Magnitude: 4.2 out of 10 — Early signal
- Stars: 113 total · +4 stars measured 2026-10-20, 11:40–13:11 UTC, ≈ 20 by evening
- Star trust: star growth looks organic
- Category: AI agents · Language: JavaScript · License: MIT · Created: 2026-10-18 · Last push: 2026-10-20
- GitHub: https://github.com/ZiChuanLan/pi-purge · Page: https://gitnova.dev/en/r/ZiChuanLan/pi-purge

## Useful for

- Test instruction-following without pi's system prompt in strict mode
- Compare model behavior across append, replace, and strict modes on one request
- Probe prompt-injection robustness when swapping the system prompt

## Why it’s here

- Star-counter measurements on 2026-10-20 (UTC), 11:40–13:11: 109 → 113 stars (+4). This is the change over that interval.
- Estimated end-of-day forecast: about +20 stars, using observed gains and the previous day.
- The repository is 2 days old and already has 113 stars. With less than two weeks of history, there's no usual pace to compare the spike against yet.
- Top new repositories this week: #192.
- About 13 forks a day — people are taking the code.

## Star trust

Star growth looks organic.

Star-trust labels are heuristics based on the repository’s behavior, not a check of every stargazer.

## Numbers

- Forks: 25
- Issues and pull requests: 1
- Watchers: 0
- Average over the last week: 30 per day
- Usual pace: too little history (under two weeks)
- Stars in the last hour (measured): 3

## Stars per day, last 3 days (oldest → newest, today is partial)

2026-10-18 … 2026-10-20: 0, 71, 4

## Spotted in now

- Top new repositories this week: #192

## More in this category

1. **google-research/harness2** — 9.7 · Breakout · AI agents · Python · +2,188 stars measured 2026-10-20, 00:22–13:08 UTC, ≈ 4,077 by evening
   A framework for recursive optimization of AI agent harnesses (prompts, tools, memory, skills) without fine-tuning model weights, with JobBench, LAB and WorkBuddy-Bench benchmarks.
   Full card: https://gitnova.dev/en/r/google-research/harness2.md
2. **MadsLorentzen/ai-job-search** — 9.6 · Breakout · AI agents · Python · +2,057 stars measured 2026-10-20, 00:22–13:08 UTC, ≈ 3,738 by evening
   A job application framework built on Claude Code: evaluates postings, tailors CVs, and writes cover letters. Runs locally, focused on the Danish job market.
   Full card: https://gitnova.dev/en/r/MadsLorentzen/ai-job-search.md
3. **anomalyco/opencode** — 7.8 · Breakout · AI agents · TypeScript · +635 stars measured 2026-10-20, 00:22–13:08 UTC, ≈ 1,161 by evening
   Open source AI coding agent with terminal and desktop interfaces. Ships with a build agent (full edit and command access) and a read-only plan agent for code analysis.
   Full card: https://gitnova.dev/en/r/anomalyco/opencode.md
4. **IsuruMaduranga/harness-engineering-101** — 7.3 · Early signal · AI agents · Python · +109 stars measured 2026-10-20, 00:22–13:08 UTC, ≈ 218 by evening
   A book series explaining how to build AI agents from first principles, treating the LLM as a stateless brain and the harness as the program that manages the conversation array. It grows a ~300-line toy agent in plain Python across 16…
   Full card: https://gitnova.dev/en/r/IsuruMaduranga/harness-engineering-101.md
5. **JoasASantos/REA-Prompts** — 7.1 · Early signal · AI agents · +383 stars measured 2026-10-20, 05:55–13:08 UTC, ≈ 585 by evening
   A curated collection of prompts for using REA (Reverse Engineer Anything) with AI coding agents to inspect binaries, firmware, kernels, games, APKs, Electron apps and websites. Covers installation, feature tracing, cloning, vulnerability…
   Full card: https://gitnova.dev/en/r/JoasASantos/REA-Prompts.md

---

Magnitude (0–10) measures how fast and how unusually interest in a repository is growing right now. It is not a quality score.
Days are UTC. “So far today” is a fact; “expected by the end of the day” is a forecast.
Summaries and use cases are written by an LLM (DeepSeek V4.1 Flash) from the README and may be inaccurate: verify specific claims (benchmarks, speed, hardware) in the repository itself.
Data as of 2026-10-20 13:25 UTC, updated every 30 minutes.