mattpocock/skills v1.1: does /wayfinder beat grill-with-docs?

Matt Pocock (AI Hero) · email · 2026-07-08 source ↗ Reader 🖥️

Agent Planning Artifacts Agent Skills Agentic Engineering Agentic SDLC Claude Code Code Quality Coding Agents Right-Sizing Boundaries & Primitives Spec-Driven Development

Pocock announces that v1.1 of his Claude Code skills repo now composes a complete agentic coding lifecycle (Grill -> Spec -> Tickets -> Implement -> Code Review), whose headline addition, /wayfinder, decomposes a plan too big for one session into agent-sized chunks mapped as GitHub issues with blocking edges so the work can be parallelized.

practical · product-update / changelog announcement (email newsletter) — A thin release-notes email for v1.1 of Pocock's Claude Code 'skills' repo: a seven-item bullet list of changes plus one paragraph on the new /wayfinder skill. The reasoning, demonstration, and any actual head-to-head comparison ('does /wayfinder beat grill-with-docs?') live behind a 'Watch the video ->' link and are not present in the text. Extraction is deliberately proportionate: descriptive product facts are marked explicit; the normative 'should' claims are reconstructions of Pocock's underlying position and marked implicit.

Structure

Part	Role
Changelog bullet list (seven items)	Enumerates the v1.1 changes: two renames, /wayfinder, Fowler-driven /code-review, /research + /prototype, grilling fixes, and the lifecycle flow
The Grill -> Spec -> Tickets -> Implement -> Code Review lifecycle	The organizing architecture the individual skills now snap into; presented as the point of the release
Wayfinder spotlight paragraph + 'planning my next course' anecdote	Elevates one skill as 'the real star' and gives an anecdotal usage claim
'Watch the video ->' link	Off-loads all argument, evidence, and the title's promised comparison to an un-transcribed video

Key terms

skill Agent Skills
A reusable, markdown-packaged, slash-invoked workflow (e.g. /wayfinder, /to-spec, /code-review) shipped in his Claude Code 'skills' repo that encodes a step of his development process as a portable, agent-runnable procedure.
/wayfinder
The new headline skill that takes an oversized plan and breaks it into 'agent-sized chunks with blocking relationships', mapping them as GitHub issues with blocking edges so work can be parallelized across sessions.
grilling
An adversarial interrogation step in which the agent interviews the human to reach alignment before building; v1.1 fixes it so the agent no longer 'self-grills' and no longer jumps straight to implementation.
agent-sized chunks
Units of work scoped small enough for a single agent/session to execute coherently — the right-sized pieces /wayfinder decomposes a large plan into.
blocking relationships / blocking edges
Declared dependency ordering between the decomposed issues (issue B is blocked by issue A), forming a dependency graph that tells the orchestration which chunks may run in parallel and which must wait.
spec vs PRD (/to-spec, /to-tickets)
A deliberate rename: '/to-prd' becomes '/to-spec' because 'the name now matches what we're actually building', and '/to-issues' becomes '/to-tickets' as 'unified terminology that's not GitHub-biased' — favouring a tool-agnostic vocabulary over PRD/GitHub-specific labels.

Propositions

explicitv1.1 renames /to-prd to /to-spec and /to-issues to /to-tickets so the command names match what is actually being built and use terminology that is not GitHub-biased.conf: high
explicit/wayfinder breaks a plan that is too big for one session into agent-sized chunks with blocking relationships, mapping them as GitHub issues with blocking edges so the work can be parallelized.conf: high
extends[FULL WORKSHOP] AI Coding For Real Engineers - Matt Pocock, AI Hero · p23 — Same author. /wayfinder is the concrete tool operationalizing the workshop's principle that parallel agent work requires a DAG of independently grabbable issues plus orchestrator/merger machinery, and that a single sequential plan caps you at one agent.
explicitdeveloped/code-review now uses Martin Fowler's refactoring smells to catch quality issues.conf: high
extendsAgentic Code Review · p14 — Adds a specific mechanism (Martin Fowler's named refactoring smells) to the broader push for structured, quality-catching AI code review beyond ad hoc judgement.
extends ←Addy Osmani's Post (Agent Skills 0.6.3 release) · p3 — Both agent-skill-pack projects independently moved their code-review step from an unranked, ad hoc check toward a codified taxonomy in the same period (Pocock's /code-review adopting Fowler's refactoring smells; this release adding severity labels and a Structural Remedies catalog) -- convergent evidence that 'code review' is becoming a formally specified lifecycle stage across competing agent-skill frameworks, not just a convention.
explicitThe grilling step was fixed so the agent no longer self-grills and no longer jumps straight to implementation.conf: high
extendsIt Ain't Broke: Why Software Fundamentals Matter More Than Ever · p6 — Continues the same 'Grill Me' lineage (adversarial interview beats default plan mode); v1.1 refines the grilling step (no self-grilling, no jumping to implementation).
supports[FULL WORKSHOP] AI Coding For Real Engineers - Matt Pocock, AI Hero · p4 — Reinforces the workshop claim that alignment is achieved by the grill-me adversarial dialogue, not by writing better plans — the fixes protect that alignment step from short-circuiting into implementation.
explicitThe skills now compose a complete lifecycle: Grill -> Spec -> Tickets -> Implement -> Code Review.conf: high
supports[FULL WORKSHOP] AI Coding For Real Engineers - Matt Pocock, AI Hero · p17 — Mapping the plan to GitHub issues (Tickets stage of the lifecycle) aligns with the workshop's stance that plans should live in an issue tracker as a visual record rather than rotting in-repo.
implicitOversized plans should be decomposed into parallelizable, agent-sized units with explicit blocking edges, rather than executed as one large sequential plan.conf: medium
supportsAI Is Changing Software Engineering. The Fundamentals Still Matter. · p2 — 'Agent-sized chunks' is an instance of investing in right-sized primitives / chunking context to the right size when working with AI.
implicitCode review should be driven by named, catalogued refactoring smells rather than ad hoc reviewer judgement.conf: medium
implicitSkill terminology should be tool-agnostic and should name the artifact actually produced (spec, ticket) rather than borrow product-management or GitHub-specific labels (PRD, issue).conf: medium

Arguments

A plan too big for one session should be decomposed into agent-sized issues with blocking edges and parallelized, rather than run as a single sequential plan. strength: low

Explicit premises:

A plan can be too big for one session ('When your plan is too big for one session...').

Implicit premises:

A single sequential plan can only be executed by one agent at a time, so it caps throughput.
Independent, agent-sized chunks can be worked in parallel across sessions and agents.
Blocking edges preserve the necessary ordering constraints, so parallelization does not sacrifice correctness of dependency order.

Inferences:

Therefore mapping the plan to a dependency graph of issues (some parallelizable, some blocked) lets more work proceed at once while respecting order.

Narratives

'The real star? Wayfinder. I've been using it for everything - even planning my next course.'
→ Enthusiasm / credibility signal that elevates one skill above the rest — an anecdotal usage claim standing in for evidence, not a demonstration of the tool's effectiveness.

Questions addressed

✅ What is new in v1.1 of the skills repo?
Two renames (/to-spec, /to-tickets), the new /wayfinder skill, a Fowler-smell-driven /code-review, supporting /research and /prototype skills, grilling fixes, and a complete Grill -> Spec -> Tickets -> Implement -> Code Review lifecycle.
⭕ Does /wayfinder beat grill-with-docs?
Not answered in the text. The phrase 'grill-with-docs' appears only in the subject line and is never defined, mentioned, or compared anywhere in the email body; any comparison lives in the linked 'Watch the video ->' video.

Gaps

incompleteThe announcement asserts benefits (parallelizable decomposition, better code review, fixed grilling) with no evidence, benchmark, or worked example — support is an anecdote plus a 'Watch the video ->' link. Most tellingly, the comparison the subject line promises ('does /wayfinder beat grill-with-docs?') is entirely absent from the body: 'grill-with-docs' is never defined or mentioned outside the subject line, so the head-to-head is not under-evidenced but wholly missing from the text — the clearest proof that the substance lives in the un-transcribed AI Hero video.
Read next: Watch and transcribe the linked AI Hero changelog video (https://www.aihero.dev/skills/skills-changelog-v1-1-wayfinder-to-spec-to-tickets-grilling-improvements), or trial the mattpocock/skills v1.1 repo directly, to recover the actual reasoning, demos, and the /wayfinder vs grill-with-docs comparison.