tdd on tomrochette.com

Jesse Vincent

tom@tomrochette.com (Tom Rochette) — Thu, 24 Sep 2026 00:00:00 +0000

Jesse Vincent (obra) is a programmer with an open-source track record long enough to predate this blog (Request Tracker, Perl 5 pumpkins, K-9 Mail, Keyboardio) who now documents a sustained personal coding-agent practice at blog.fsck.com and builds Superpowers at Prime Radiant. Facts below verified as of 2026-09-24.

He is the practitioner who turned agentic coding from loose prompts into a codified, eval-measured methodology you can install today; the price of that discipline is tokens, speed, and a strong-opinions workflow that not every codebase or culture will accept.

What it is
#

A long-running personal blog (541 posts as of 2026-09-24) where he publishes point-in-time writeups of how he works with agents, plus Superpowers, an MIT-licensed skill suite (291k stars, 26k forks, 682 commits as of 2026-09-24) that packages his brainstorm, plan, worktree, subagent-driven development, TDD, and two-axis code review loop as mandatory workflows for 16 coding-agent harnesses. Wikipedia credits him with publishing the Architect/Implementer pattern and a SKILL.md-formatted skills approach about a week before Anthropic shipped its native skills framework. He is founder and CEO of Prime Radiant, the applied research lab behind Superpowers, which sells commercial support around it and maintains a public eval suite.

Status
#

Very active and shipping on a cadence: Superpowers 6.4 landed 2026-09-21 (adding Meta Muse, OpenCode 2.0, and Qwen Code support), a Birds-of-a-Feather on agentic engineering with Simon Willison was announced 2026-09-22 for San Francisco, and recent months brought a vibe-coded C compiler that builds SQLite (2026-08-21), Superpowers 6 (2026-06-15, 50 percent faster and 60 percent cheaper by their evals), a hiring post (2026-06-12), and 5.1.0 (2026-05-04). His writeups say users’ most common lament is that Superpowers burns many tokens and takes longer than going without it, and 6.4 concedes that bare-metal builds are faster and cheaper while arguing they are significantly buggier per his eval suite. Third-party pull is visible: HN threads recommend Superpowers as the fix for bad agent experiences, Willison cites it as the reference implementation of TDD-with-agents, OpenAI’s Codex announcements were read as bundling it, and orchestration stacks like metaswarm declare it, alongside Yegge’s beads, an essential foundation.

Strengths
#

The workflow is specific and copyable: Socratic brainstorming, plans written for “an engineer with zero context and questionable taste”, isolated worktrees, fresh subagents per task with review after each, and reviews checked on spec compliance and quality separately.
He measures his own system instead of asserting it: the public superpowers-evals repo and the 6.0/6.4 posts report speed, cost, and bug deltas across harnesses and models.
He reports failures, not just wins, from the session that tried to delete his tests to the goal loop that stopped early on the compiler project, which makes the writeups unusually trustworthy.
His blog doubles as a skills-engineering manual: pressure-testing skills against subagents, Cialdini-style persuasion principles as reliability levers, and skills mined from his own conversation history.
The ecosystem compounds around him, so adopting the suite buys you a community and a lineage (superpowers-rails, metaswarm) rather than a lone repo.

Cautions
#

Cost and speed are the known tax; if your budget cannot absorb subagent-driven development’s overhead, his own release notes tell you bare metal wins on both axes.
The methodology is mandatory by design (skills “must” be used, TDD enforced, code written before tests gets deleted), which reads as rigidity if your context is a legacy codebase or a non-TDD culture.
Behavior shifts with every frontier model release, and his posts are explicitly point-in-time snapshots, so six-month-old methodology posts need re-verification before you copy them.
The visual companion feature phones home version telemetry by default via Prime Radiant’s site, with an opt-out environment variable, which is disclosed but worth knowing.

Compared to
#

Simon Willison: Willison surveys the whole landscape daily and Vincent systematizes one practice deeply, and the two co-host the October 2026 agentic-engineering BoF in San Francisco; read Willison for breadth, Vincent for a runnable process.
Thorsten Ball: both write first-person practitioner notes, but Ball works at the harness-building layer while Vincent works at the methodology layer on top of harnesses like the ones Ball builds.
Steve Yegge: Yegge’s beads tracker and Gas Town orchestration thesis sit above Vincent’s skills layer, and real stacks (metaswarm) combine the two; choose Vincent for disciplined task execution, Yegge for tracking and multi-agent orchestration.

Bottom line
#

Recommended for engineers who want a tested, opinionated agentic workflow they can install today, and for anyone designing agent skills, since his posts are the best public record of how to engineer them. Not for token-budget-tight users, TDD-averse codebases, or readers who want minimal process between idea and diff.

Changes
#

2026-09-24 - Created.

References
#

https://blog.fsck.com/ - the blog homepage: cadence, recent posts, 541-post archive, and the Prime Radiant/Best Practical/Keyboardio work links
https://blog.fsck.com/2025/10/05/how-im-using-coding-agents-in-september-2025/ - the September 2025 methodology writeup
https://blog.fsck.com/2025/10/09/superpowers/ - the Superpowers release announcement and skills philosophy
https://blog.fsck.com/2026/06/15/Superpowers-6/ - the Superpowers 6 release: eval-driven 50/60 percent improvements and the token-cost admission
https://blog.fsck.com/2026/09/21/superpowers-6.4/ - the 6.4 release: new harness support, bare-metal comparison, diagnosing-superpowers skill
https://blog.fsck.com/2026/08/21/i-vibe-coded-a-c-compiler/ - the autonomous C-compiler experiment grounding his long-horizon loop practice
https://github.com/obra/superpowers - the repo: MIT license, 291k stars and 26k forks as of 2026-09-24, 16 documented harness installs, commercial support, and telemetry disclosure
https://en.wikipedia.org/wiki/Jesse_Vincent - biography: RT, Best Practical, Perl roles, K-9 Mail, Keyboardio, Prime Radiant CEO, and the SKILL.md-before-Anthropic timeline
https://hn.algolia.com/api/v1/search?query=jesse%20vincent%20superpowers&hitsPerPage=8 - third-party reception: HN recommendations, the Superpowers 6 thread, metaswarm’s reliance on it, and Willison citing it as the TDD reference

Kent Beck

tom@tomrochette.com (Tom Rochette) — Thu, 24 Sep 2026 00:00:00 +0000

Kent Beck, creator of Extreme Programming and test-driven development, publishes his current thinking on software design and AI-era practice on the Software Design: Tidy First? newsletter. Facts below verified as of 2026-09-24.

He is the most credible methodology voice on which parts of classic software craft survive when agents write the code, and his boundary is process and economics: he tells you how to work, not which tool or model to buy.

What it is
#

A weekly newsletter (over 126,000 subscribers as of 2026-09-24) backed by a forty-year methodology portfolio: Extreme Programming Explained, Test-Driven Development by Example, Implementation Patterns, the 2023 book Tidy First?, and co-creating JUnit with Erich Gamma. His AI arc runs from the April 2023 essay “90% of My Skills Are Now Worth $0” (the repricing thesis: “the value of 90% of my skills just dropped to $0. The leverage for the remaining 10% went up 1000x”) to June 2025’s “Augmented Coding: Beyond the Vibes”, which separates augmented coding, where you still care about code, tests, and design, from vibe coding, where you only care about behavior. He grounds that distinction in a real project, a B+ tree library in Rust and Python built with agents over about four weeks, with his full system prompt and intervention habits published in the post. Around the newsletter sit consulting, speaking (topics like “Beyond Vibes: What Augmented Coding Actually Requires”), a Still Burning podcast, and Tidy Together, the in-progress third book of his Empirical Software Design series.

Status
#

Active, and newly central to the process debate. Cadence is weekly per his own site, and the newsletter’s top-posts archive shows entries dated December 12 2025 and January 16, April 16, and May 15 (Substack omits the year on current-year posts, so those are 2026). He is independent: Wikipedia records his Facebook years (2011-2019) and a 2019 software fellow role at Gusto, but kentbeck.com now presents him as consultant, writer, artist, and speaker, with draft chapters of Tidy Together going to paid subscribers. The AI-era posts themselves drew modest Hacker News traction (his own submission of the augmented-coding essay sat at 5 points as of 2026-09-24), yet the framing keeps resurfacing in threads asking whether XP should be revisited for the agent era, which is the slower, stickier kind of influence.

Strengths
#

Authority with receipts: the person who defined TDD and XP re-deriving his own methodology for agents carries weight no commentator can match.
He experiments on real code and shows his work: the B+ tree build came with the system prompt, the time logs, and a candid verdict, correct and fast but with more accidental complexity than he likes.
He names the distinctions the industry now argues with: augmented versus vibe coding, and the three warning signs of an agent going off track (loops, unrequested features, cheating on tests).
The economics essays give leaders a decision frame instead of a prediction: Programming Deflation’s substitution-versus-Jevons pairing ends with capabilities to build that pay off under either future.
Short, quotable essays, one idea each, with no tool churn to wade through.

Cautions
#

This is a process feed, not a tracking feed: expect near zero coverage of specific tools, models, and prices, so pair it with a practitioner blog for that layer.
His evidence is personal and he says so: “I’m extrapolating wildly from a couple of experiences, which is what I do.”
Even his own commenters push back on the emotional cost: one writes that LLM-generated code “feels to me like having no soul”, a tension he does not resolve.
The best material (draft book chapters, weekly Thinkies) sits behind the paid tier.

Compared to
#

Simon Willison: the daily tool-level chronicler versus Beck’s weekly process level; use Willison to learn what changed this week, Beck to decide what it means for how your team works.
Andrej Karpathy: the model-and-training-level view, and the coiner of the “vibe coding” term Beck argues against, versus Beck’s working discipline of tests and tidiness under agents.
Steve Yegge: manifesto-scale provocations versus small falsifiable essays; read Yegge for the argument, Beck for the practice.

Bottom line
#

Recommended for senior engineers and leads who need to rework team process, review habits, and design discipline for agent-heavy development. Not for readers who want tool comparisons, model benchmarks, or news.

Changes
#

2026-09-24 - Created.

References
#

https://tidyfirst.substack.com/p/augmented-coding-beyond-the-vibes - the augmented versus vibe coding essay, the B+ tree case study, and the published system prompt
https://tidyfirst.substack.com/p/90-of-my-skills-are-now-worth-0 - the 2023 repricing thesis and the 3X Explore/Expand/Extract framing
https://tidyfirst.substack.com/p/programming-deflation - the substitution versus Jevons economics essay (2025-09-15)
https://tidyfirst.substack.com/p/canon-tdd - his definitive TDD workflow essay (2023-12-11)
https://tidyfirst.substack.com/ - the newsletter home: over 126,000 subscribers as of 2026-09-24
https://tidyfirst.substack.com/archive?sort=top - posting cadence evidence with dated entries into 2026
https://www.kentbeck.com/ - current projects (Tidy Together, augmented-coding research and speaking), subscriber and open-rate figures, consulting and podcast
https://en.wikipedia.org/wiki/Kent_Beck - the methodology record: XP, TDD, JUnit, the Agile Manifesto, Facebook and Gusto years, book list
https://hn.algolia.com/api/v1/search?query=%22augmented%20coding%22%20kent%20beck&hitsPerPage=8 - third-party HN evidence, including his own submission at 5 points and citations in XP-revision threads (as of 2026-09-24)

tdd on tomrochette.com

Jesse Vincent

What it is
#

Status
#

Strengths
#

Cautions
#

Compared to
#

Bottom line
#

Top 5 recommended reading
#

Changes
#

See also
#

References
#

Kent Beck

What it is
#

Status
#

Strengths
#

Cautions
#

Compared to
#

Bottom line
#

Top 5 recommended reading
#

Changes
#

See also
#

References
#

tdd on tomrochette.com

Jesse Vincent

What it is #

Status #

Strengths #

Cautions #

Compared to #

Bottom line #

Top 5 recommended reading #

Changes #

See also #

References #

Kent Beck

What it is #

Status #

Strengths #

Cautions #

Compared to #

Bottom line #

Top 5 recommended reading #

Changes #

See also #

References #

What it is
#

Status
#

Strengths
#

Cautions
#

Compared to
#

Bottom line
#

Top 5 recommended reading
#

Changes
#

See also
#

References
#

What it is
#

Status
#

Strengths
#

Cautions
#

Compared to
#

Bottom line
#

Top 5 recommended reading
#

Changes
#

See also
#

References
#