Pi You Should Know — Development skill for Claude Code
Pi extension: a port of Claude Code's you-should-know mod; every few steps, ask a fork of the conversation whether there is something you should know but probably missed.
How to install Pi You Should Know
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open aliceisjustplaying/pi-you-should-know and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Pi You Should Know does
Pi extension: a port of Claude Code's you-should-know mod; every few steps, ask a fork of the conversation whether there is something you should know but probably missed.
Alternatives in Development
- Param Discover — Discover hidden HTTP parameters on a URL or list of URLs using Arjun (or x8 fallback) 4.5k ★
- Haft — Engineering decisions engine that know when they're stale 1.4k ★
- Claude Code Everything You Need To Know — Ultimate all-in-one guide to mastering Claude Code 1.3k ★
README
pi-you-should-know
A Pi extension that occasionally taps you on the shoulder: "heads up, you probably missed this."
While an agent works, a second copy of the conversation is asked one question every few steps: is there something the user should know but probably didn't notice? Usually the answer is no and you see nothing. When the answer is yes, a short note appears above the editor, with a pointer to where it came from (a file and line, a test, a command).
It's a port of the "you should know" plugin built into Claude Code.
Using it
- A note shows up above the editor, like
✦ Heads up · The agent loosened the test threshold from 0.8 to 0.65. /yskor alt+shift+y: open the top note. You can learn more about it, mark it as something you knew, send it to the agent or dismiss it.- alt+x: dismiss the top note.
/ysk off,/ysk on,/ysk status: turn it off and on, or see how many checks it has run.
If you keep ignoring notes, it checks less often. Answering any note resets that.
In RPC clients
Without Pi's terminal UI (for example in T3 Code), each note is sent once as a notification in the form `[ysk:] · ()`, followed by its explanation, so the client can show and answer it. A note found after a run has finished is sent at the start of the next run. `/ysk` still works there through the client's dialogs.
With herdr subagents
When the extension runs inside a [pi-herdr-subagents](https://github.com/aliceisjustplaying/pi-herdr-subagents) subagent, its notes aren't shown in the subagent's pane. They're passed to the parent agent as a message instead, labeled with the subagent's name, so the agent that's coordinating the work can act on them.
Cost
Each check reuses the conversation the main agent already cached, so it costs a fraction of a normal turn. Every check is logged, with token usage, to `~/.pi/agent/you-should-know/checks.jsonl`.
Install
pi install git:github.com/aliceisjustplaying/pi-you-should
Related Skills
Pantheon Insurgents Advantage
Invoke when outgunned, outresourced, or outranked in a contest you need to win — or when you have built someth
Claude Mod Skill Suggester
Suggests one of your skills from what you are working on: your message, the last few turns, the files Claude r
Agent Conv CLI
Read and continue your local Claude Code / Codex CLI / Cursor conversation history from the terminal (chats/re
Claude Code Usage Quota Mod
Know your Claude limits before they hit you (Live band above the Claude Code prompt): 5 Hour and Weekly usage
Skill-porter - Gemini CLI Extension
Converts Claude Code skills to Gemini CLI extensions and vice versa. Use when the user wants to make a skill c
Engineer Capability Promote
Decide whether recurring work belongs in reuse, a project rule, an Engineer module, a skill extension, or a ne
Related Agents
Odd Explorer
Read-only code explorer for ODD. Use to map code, trace a flow, or find where something lives when answering i
Part 13: Memory Bridge (Give Coding Agents Your Brain)
When you spawn Codex or Claude Code to build something, they start blind. They don't know your architecture de
Researcher Architect
Investigates how existing NGM functionality works, locates relevant files, and produces technical designs for