A question-and-answer planning process that interviews the user in rounds before work begins. It explores business needs and, depending on the chosen path, technical and implementation details.

It helps produce a design tree, acceptance plan, architecture notes, and implementation tasks for a feature.

It reduces misunderstandings by settling important decisions with the user before the coding work starts.

grill — Skill for Claude Code and Codex by ntorga

Install

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.

Any agent

npx skills add ntorga/agent-starter-kit --skill grill

Clone the repo

git clone --depth 1 https://github.com/ntorga/agent-starter-kit

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/ntorga/agent-starter-kit/grill/github.svg)](https://agentmods.dev/skills/ntorga/agent-starter-kit/grill)

Your own site

<a href="https://agentmods.dev/skills/ntorga/agent-starter-kit/grill"><img src="https://agentmods.dev/badge/skills/ntorga/agent-starter-kit/grill/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/ntorga/agent-starter-kit/grill"><img src="https://agentmods.dev/badge/skills/ntorga/agent-starter-kit/grill.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 33 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,174 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What 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 5.1	$0.00033	$0.01174
Opus 5	$0.00016	$0.00587
Sonnet 5	$0.00007	$0.00235
Haiku 4.5	$0.00003	$0.00117

Measured 2d ago against content hash e5a46503b319, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-14, from the pricing page.

Security

Grade A, and why

grill 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.

skills/grill/SKILL.md · 69 lines

How it starts

The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Purpose

Reach shared understanding with the user before execution. The Architect owns the design tree — the decisions the feature must settle, mapped with dependencies, pruned by path, ranked by impact. You work the frontier in rounds: ask up to 5 frontier questions, wait for the user's answers, transcribe the settled decisions into the artifacts. After the user confirms, execution is autonomous. The user is bothered in rounds, not on every micro-delivery.

Procedure

The path determines which branches the tree explores. Step 1 resolves it from long-term memory, or asks the user when none is recorded:

Beginner — business questions only. The Architect infers architecture from the maintenance, lightweight, and safe principles (KISS, single responsibility, safe boundaries — rules/code/general.md). Ceiling: 30 business questions.
Tinkerer — business and architecture questions. Ceiling: 30 business, 10 architecture.
Pro — business, architecture, and implementation questions. Ceiling: 100 questions per branch.

Artifacts live at .memory/plan/<feature-slug>/:

tree.md — the design tree. The Architect owns it.
plan.md — acceptance criteria, from the settled business decisions.
arch.md — directory structure, layer separation, reference projects. Tinkerer and pro. Beginner gets defaults.
impl.md — epics grouped at one day of work, with parallelizable groups.

You transcribe plan.md, arch.md, and impl.md as decisions settle. The Architect owns tree.md and the grounded impl.md.

Select path. Check long-term memory (uses: skills/agent-memory/SKILL.md) for a recorded grill path under ## Preferences. If one exists, proceed with it — state the path in the first message to the user so they can override in one word. If the user overrides, replace the recorded entry. If none exists, ask the user to choose: beginner, tinkerer, or pro. Record the choice under ## Preferences as [grill] path: <beginner | tinkerer | pro>.
Build the tree. Dispatch the Architect (personas/architect.md) (follows: skills/dispatch/SKILL.md) in tree mode with the selected path. The Architect writes tree.md.
Work the frontier in rounds. The frontier is every open node in tree.md whose dependencies are all settled. Send up to 5 frontier questions per round, ordered by impact. Each question carries the tree's recommendation. Wait for the user's answers before the next round.

Read the full file on GitHub · 69 lines

Changes

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 · 69 lines · 33 tokens per session scan A e5a46503b319

Subscribe to this mod's changes

grill is a skill published in the GitHub repository ntorga/agent-starter-kit (142 stars, last pushed 2d ago), licensed MIT. It adds 33 tokens to every session and 1,174 once invoked, about $0.0002 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 2026-09-13.

Multi-Agent Planning agent-framework ai-agents coding-assistant llm-orchestration model-agnostic

agent-framework-py-release

Use when cutting a Python release for the microsoft/agent-framework monorepo. Triggers on "bump py versions", "cut a python release", "prepare release PR for python", "release py packages", "bump python to X.Y.Z", or similar requests to bump Python package versions and prepare a release PR. Handles all four lifecycle…

microsoft/agent-framework · 103 tokens

python-package-management

Guide for managing packages in the Agent Framework Python monorepo, including creating new connector packages, versioning, and the lazy-loading pattern. Use this when adding, modifying, or releasing packages.

microsoft/agent-framework · 43 tokens

foundry-hosted-agent-validation

Step-by-step process for validating a Python Foundry hosted agent sample (under python/samples/04-hosting/foundry-hosted-agents/) end to end — running it locally (native runtime and azd ai agent run) and after deploying it to an Azure AI Foundry project with azd. Use this when asked to validate a hosted agent sample.

microsoft/agent-framework · 82 tokens

grill