Skip to content
SkillModel-led403 tokens in use

Grilling

Interview the user in numbered rounds, each question with a recommended answer, until every branch of the design tree is settled.

328,791 stars on the repository MIT Licensemattpocock/skills · skills/productivity/grilling
claude plugin install mattpocock-skills@claude-plugins-official
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing Grilling costs

38tokens idle
403tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

Nothing detected: no shell, network, file writes, git or credentials in its files.

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Fair
  • Frontmatter12/15
  • Structure3/25
  • Examples6/15
  • Output15/15
  • Size15/15
  • Resources0/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent

No findings.

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Excellent
  • Development activity96
  • Maintenance100
  • Community56
  • Maturity32
  • Momentum40

Grilling is the reusable interview method behind the grill command. The agent maps a plan as a design tree, finds the frontier of decisions whose prerequisites are settled, and asks all of them in one numbered round with a recommended answer for each.

It looks up facts itself through sub-agents and asks you only for decisions.

What You Get

  • A numbered list of questions per round, each with a recommended answer
  • Questions worded so that yes accepts the recommendation
  • Facts gathered by sub-agents with fresh context
  • A shared understanding confirmed before any action

Common Use Cases

  • Stress-testing a plan before building
  • Settling a set of linked design decisions in few rounds
  • Interviewing yourself about an idea with an agent that does the research
  • Feeding a clear brief into later steps

What Makes It Unique

It asks the whole frontier at once, so each round holds every ready question, which shortens long interviews. Each question carries a recommended answer, and the agent never asks you for facts it could find itself.

Under The Hood

The skill is one Markdown file plus an agent metadata file for Codex. It tells the agent to dispatch sub-agents with a fresh context window for factual lookups such as filesystem or tool checks. It ships no scripts and runs no commands of its own.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers