TGViewer
Channel Public Channel
Matt Pocock

Matt Pocock

@mattpocockuk

I teach devs for a living. Author of Total TypeScript and Al Hero. Ex-@vercel. Used to be a voice coach.
Open in Telegram
Subscribers
16
Photos
6
Videos
2
Links
21
Recent Posts 20 shown
Post #83 82
One clarification for folks using /wayfinder:

The flow for big work should be:

/wayfinder -> /to-spec -> /to-tickets -> /implement

Once the /wayfinder map is complete, you turn it into a spec.

Some folks are using /wayfinder as the ENTIRE flow - from grilling to prototyping to shipped work

It certainly can be used that way - I've been doing that for non-coding stuff like course creation.

But for coding I much prefer creating a spec and handing off implementation to an AFK agent. Means I can focus on other things for a bit while it churns away.

I'll be putting all of this in an upcoming tutorial. Thanks for all the great feedback on v1.1. v1.2 is in the works and looking good.
  • ❤ 1
Post #82 77
Matt Pocock Thinking about a workflow like this for helping prevent comprehension debt on a fast-moving repo: Fast-moving repo w/lots of changes -> once per day/week, grab a diff of the changes -> feed it to an LLM to output a podcast transcript, focusing on the 'why'…
I don't really like the "extra-curricular" aspect of this. It works for me because I choose my working hours, but a video briefing might work better.
Post #81 75
Thinking about a workflow like this for helping prevent comprehension debt on a fast-moving repo:

Fast-moving repo w/lots of changes
-> once per day/week, grab a diff of the changes
-> feed it to an LLM to output a podcast transcript, focusing on the 'why' of what changed more than the 'what
-> feed the podcast transcript to a TTS API
-> publish to a private podcast feed
-> listen on your commute/morning run/lunch break

I'm doing this for my personal wiki, which ingests from X/Slack/Discord/Gmail, and it's super-efficient for keeping me in the loop.
Post #80 69
A junior just asked me "what do I do if I'm being grilled about something I don't understand?"

Try this:

1. In a grilling session
2. "I don't understand this"
3. "/handoff to a teaching agent to understand this"
4. Pick up the handoff doc with the /teach skill, in your teaching workspace
5. Learn it
6. Go back to the grilling session and pick up where you left off

Keeps the grilling session clean, records progress in your /teach skill, and helps you learn
Post #79 64
Someone DM'd me saying they're using /grill-me in technical interviews

Watch the candidate work. See how they answer questions and push back. See whether the AI drives them, or they drive the AI.

Honestly genius
  • 👍 1
Post #78 63
Matt Pocock 1. Delete the docs you create to explain your code 2. Take the tokens you save on updating those docs 3. Spend them on making your code self-explanatory
The only exception are things that the code cannot explain itself, like alternatives considered or domain language. I.e. ADR's and glossaries
Post #77 62
1. Delete the docs you create to explain your code

2. Take the tokens you save on updating those docs

3. Spend them on making your code self-explanatory
Post #76 59
Reworded /grill-me so it can now truly be used anywhere - not just engineering.

"design tree" -> "decision tree"
"every aspect of this plan" -> "every aspect of this"
"codebase" -> "environment"
"Do not enact the plan" -> "Do not act on it"

https://github.com/mattpocock/skills/pull/532
GitHub grilling: reword the primitive for general use by mattpocock · Pull Request #532 · mattpocock/skills Reword the grilling primitive so it reads as a stress-test of any plan, decision, or idea — not only a software build plan. grill-me is left untouched; the behaviour change lives in the shared prim...
Post #75 69
I feel like I've developed a clear rationale for when to /compact, when to /clear, and when to /handoff

But I'm having trouble articulating it.

So, what's your mental model for when to do each?
Post #74 69
Proposal: a /research skill

It's really simple - just spins up a background agent to look at high-trust sources, and saves them in a markdown file.

Worth keeping? Or pure no-op?


https://github.com/mattpocock/skills/pull/409
GitHub Add the research skill by mattpocock · Pull Request #409 · mattpocock/skills Adds research — a small, model-invoked engineering skill. What it does Spins up a background agent to investigate a question against primary sources (official docs, source code, specs, first-party ...
Post #73 63
Matt Pocock I'm working on a new skill which helps you plan enormous chunks of work, far larger than /grill-me can It identifies the frontier of decisions and where the fog of war is It suggests prototyping, research or grilling depending on what the decision is As…
Or /wayfinder
Post #72 64
I'm working on a new skill which helps you plan enormous chunks of work, far larger than /grill-me can

It identifies the frontier of decisions and where the fog of war is It suggests prototyping, research or grilling depending on what the decision is

As you push back the fog of war it reveals more decisions

It scopes down every decision so it fits in a separate context window (for maximum parallelisation)

But I need to give it a name. It's lamely called /decision-mapping right now.

I was thinking /pathfinder:

/pathfinder I want to build an enormous greenfield app

WDYT?
Post #71 59
/to-prd is changing to /to-spec
/to-issues is changing to /to-tickets

Apologies for the churn, but I've been wanting to make this change for a while and now's the time to do it.
Post #70
Matt Pocock pinned «My last AI Coding Cohort was the most successful course I've ever released. 2,500+ students worked with Claude Code for 2 weeks, building a real app with AFK agents and software fundamentals. So, we made version 2: - Use any coding agent you like - Updates…»
Post #69 90
Another layer of documentation I'm considering (along with CONTEXT.md and ADR's) is a list of all the agreed test seams in the app

Agents simply cannot be trusted to make good decisions about what to test, and at what seam.

For every small change, they extract out only what they've built into a testable function and test that.

It leads to a patchwork nightmare of tests that break as soon as the implementation changes.
Post #68 89
My last AI Coding Cohort was the most successful course I've ever released.

2,500+ students worked with Claude Code for 2 weeks, building a real app with AFK agents and software fundamentals.

So, we made version 2:

- Use any coding agent you like
- Updates for every skill
- Uses Sandcastle for AFK agents

Feeling behind the curve? How would you like to be... ahead of it?

Starts June 1st. Get your seat now:

https://www.aihero.dev/cohorts/ai-coding-for-real-engineers-m0k0w
www.aihero.dev AI Coding for Real Engineers A two-week cohort for developers who want to use AI coding tools on real production-style work: context gathering, planning, steering, feedback loops, AFK agents, and human-in-the-loop review.
Post #67 68
Long skills are such a red flag to me

- Hard to audit (and therefore, trust)
- Hard to edit (more text, harder to maintain)
- Expensive to run (more text, more tokens)

The shorter the skill, the better IMO
Post #65 135
I'm in a mood to write some library code

What things in the AI SDK do you wish were easier?
Post #64 161
Here's the difference between an eval and a unit test:

unitTest: () => 'pass' | 'fail'
eval: () => {
// A score between 0-100
score: number
}

Evals are for cases where unit tests won't cut it: readability, relevance, humor, factuality.
Older posts →

About this channel

How can I read @mattpocockuk without a Telegram account?
TGViewer shows the public web preview Telegram publishes for Matt Pocock: recent posts, photos, videos and the subscriber count, with no app, login or account.
How many subscribers does Matt Pocock have?
Matt Pocock (@mattpocockuk) has 16 subscribers on Telegram, refreshed roughly every 30 minutes.
Does Matt Pocock know I viewed it here?
No. Public channel previews carry no viewer identity, and TGViewer has no accounts or tracking of what you look up.
Threads Profile ViewerView any public Threads profile without an account.Open ThreadLook →Writing with AI? Make it sound human.Metric37 rewrites AI drafts so they read naturally. Free AI detector, 1,500 words free.Try Metric37 →