Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add rules/tonyghiani/ai-essentials/numbered-repliesgit clone --depth 1 https://github.com/tonyghiani/ai-essentialsWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00478 | $0.00478 |
| Opus 5 | $0.00239 | $0.00239 |
| Sonnet 5 | $0.00096 | $0.00096 |
| Haiku 4.5 | $0.00048 | $0.00048 |
Grade A, and why
numbered-replies scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Numbered Replies
When you ask the user a numbered list of questions, expect a numbered reply where each number corresponds to your question of the same number.
How to interpret
- A reply starting with
1.(or1)or just a list of bullets in order) maps to your questions by index. - Do not ask the user to disambiguate which answer goes with which question — the index does that.
- Short answers are intentional and complete:
Out of scope— that aspect is not being addressedSame flag— keep the existing flag, no changeboroption B— picked the lettered choice you offeredkeep it/drop it/as-is— accept whatever defaults you proposed
- Missing numbers mean "no change / skip". Don't re-ask.
- If the user reorders items, match by content not by position. If ambiguous, ask one targeted question.
How to ask questions to invite this format
When you have multiple questions, present them as a numbered list. Keep each question short and answerable in one phrase:
1. Should we keep behavior X or change to Y?
2. What's the timeout — 30s, 60s, or other?
3. Edge case: empty input — error, default, or skip?
Not this:
I have a few things I'd like to ask you about. First, regarding behavior X,
should we keep it or change it to Y? Then there's the question of timeout,
which I'm thinking should be 30 seconds but could be 60 — what do you think?
Also, what should we do for empty input?
The numbered format lets the user answer in one short turn instead of writing a paragraph.
When NOT to use this format
- Single question — just ask it directly.
- Open-ended exploration ("what do you think about X?") — let the user answer freely.
- Yes/no on a single decision — use a structured multiple-choice prompt if your tool supports one; otherwise ask directly.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 48 lines · 478 tokens per session scan A a5578edd769b
numbered-replies is a cursor rule published in the GitHub repository tonyghiani/ai-essentials (6 stars, last pushed 24d ago), licensed MIT. It adds 478 tokens to every session, about $0.0024 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other cursor rules, from other repositories
050-plan
When the user types /plan or asks to create a project plan, feature PRD, or retrospective.
dreamd-recall
Recall lessons, decisions, and prior context from the .agent/ memory daemon. Use when starting work in a project that has a .agent/ folder, when the user references a past decision, or when you are about to make a choice that has a documented prior.
session-memory
Use at conversation wrap-up or when the user explicitly indicates end-of-session — capture residual lessons not captured in-flight.
context-recorder-system
Context Recorder System (记录员系统) - 模块化索引文件.
rules
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
misc-documenting-learnings-and-clarifying-assumptions
Documenting Learnings and Clarifying Assumptions for Efficient Task Execution.