Skip to content
SkillModel-led2,975 tokens in use

Ask Matt

A router over Matt Pocock's engineering skills that tells you which skill or flow fits your situation, from idea to shipped code.

328,754 stars on the repository MIT Licensemattpocock/skills · skills/engineering/ask-matt
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 Ask Matt costs

22tokens idle
2,975tokens in use
3,931tokens, up to

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Up to adds the files its instructions link. 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 and code blocks. 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 →
Good
  • Frontmatter8/15
  • Structure19/25
  • Examples3/15
  • Output10/15
  • Size15/15
  • Resources7/15

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

Ask Matt answers the question of which skill to reach for. It maps the repository’s skills into one main flow from idea to ship, two on-ramps (incoming issues and broken code), a codebase health step, a vocabulary layer and a set of standalone tools.

You invoke it yourself. Before it states what a skill does or recommends skipping a step, it reads that skill’s SKILL.md.

What You Get

  • A main flow from idea to ship: grill-with-docs, an optional prototype detour, to-spec and to-tickets for multi-session builds, then implement or implement-spec, code-review and retro
  • On-ramps for triage, diagnosing-bugs and wayfinder that merge onto the main flow
  • Guidance on context hygiene: when to continue, clear, hand off, use a subagent or compact at a phase boundary
  • Short descriptions of standalone skills such as grill-me, prototype, research, wizard, wait-what and teach

Common Use Cases

  • Starting a feature and not knowing which skill to run first
  • Deciding between implement for each ticket and implement-spec for a whole spec
  • Choosing what to do with the context window between the phases of a session

What Makes It Unique

Most skill collections list their skills. Ask Matt lays them out as flows with branch points, such as whether a question needs a runnable prototype or whether a build spans several sessions, and it names the order to run them in. It also points to PHASE-BOUNDARIES.md, a five-question tree for choosing between continue, clear, handoff, subagent and compact.

Under The Hood

The skill is one Markdown document that the model reads when you invoke it by name; automatic model invocation is disabled in its frontmatter. It runs no scripts and calls no services. It links to PHASE-BOUNDARIES.md for the phase-boundary decision tree, and it tells the model to read each recommended skill’s SKILL.md before describing it.

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