C9 · Week 10: Crit the critics
No new prototype this week — the write-up is the artefact, and it feeds straight into the portfolio.
find a high-profile agentic-coding thinker who has pontificated on the best way to use these tools — blog posts are fine — and read their advice against your own practice: what tracks, what’s nonsense, and the examples that prove it
By now your practice is on the record: a process overview and a reflection for every crit so far, and an in-flight A3. This week you read somebody else’s doctrine against it. Pick a voice with reach — someone who has published a position on how this work should be done: a manifesto, a named workflow with a school of thought behind it, a running blog of pronouncements. The field is not short of them.
Then crit the advice the way the studio crits your prototypes: specifically. Where does your own record back the pundit up, and where would following them have cost you? Agreeing everywhere is as thin as sneering everywhere; a real read comes back with a mixed verdict and the receipts for it. Take a position; “it depends” is not a verdict.
Someone else’s claims set against your own evidence is exactly the shape of the portfolio’s workflow justification, and the receipts are evidence of process you’ve already banked. And your A3 keeps moving all the while — still deployed, still green at the cutoff.
The spec
The brief on this page poses the problem and leaves room for your response. The spec is the fixed contract: what your tutor considers when judging whether that response meets the requirements. Some lines can be checked mechanically; the rest call for human judgement.
- a named thinker, picked from the wild: high-profile, with a published position on how to use agentic coding tools — blog posts are fine
- the response exists: which parts of their advice track and which are nonsense, argued in your own voice
- the examples are real: specific claims of theirs set against specific moments from your own work — commits, transcripts, reflections
- your final project is still deployed and green at the cutoff — the reading didn't stall the build
- the repo shows the process — commits that grew with the work, a process overview in PROCESS.md, and the week's reflection in reflections/crit-9.md
- you can account for how you directed, grounded and corrected the work