CtrlK
BlogDocsLog inGet started
Tessl Logo

ainativedev/latest-aidevcon-speakers-london-2026

AI Native DevCon 2026 London — all conference sessions as interactive skills

66

Quality

83%

Does it follow best practices?

Impact

No eval scenarios have been run

SecuritybySnyk

Risky

Do not use without reviewing

Overview
Quality
Evals
Security
Files

quotes.mdtalk-maple-harness-engineering/

Notable verbatim quotes

#QuoteTopicSection
1"harness engineering, which is a thing that is kind of near and dear to my heart kind of having invented the term here"Origin / framingIntro
2"I had an insane idea that I would try and get this tool to do my job."Origin storyOrigin story
3"if you're not in the habit of kind of completely retooling your stack in the way you work with every point release of the model, you are in a way missing out on what it is that you can achieve with these tools"Pace of disruptionPace of disruption
4"Human time is the fundamentally scarce resource that we have."Foundational constraintsNew constraints
5"we have to find ways to make all these nonfunctional requirements of writing good software legible to the agent"Write it downWriting down
6"I never want to give the same review feedback twice."Write-it-down ratchetReact/suspense analogy
7"Harness engineering is making context around what it means to do a good job legible. And then just in time surface to the agent over the course of its trajectories in order to steer and refine its output to make sure that every PR we get adheres to the golden thread of what we consider to be acceptable, high quality aligned software."DefinitionDefining harness engineering
8"I try and put my interventions as far right in the process as I can. In order to minimize my own synchronous time having to engage with these issues."Shift rightShift right, not left
9"it's up to us to prune latent space, to tell it which choices we want to make"Pruning latent spacePruning latent space
10"all the code in the repository of itself outside of the documentation knowledge base is also prompts"Code-as-promptCode-as-prompt
11"the most important thing that ends up in that agents.md Is a numbered set of steps that we expect the model to go through over every rollout that we do, over every session"agents.md structurePhase 1
12"they respond really well to descriptive error messages that point them to run books for remediation steps"Just-in-time injectionPhase 2
13"we can absolutely vibe a set of guardrails into place. With 100% code coverage and exhausted table driven tests."Building guardrailsPhase 2
14"We don't want to jam a ton of rules in here. Because we're going to chop up latent space too much."agents.md as mapagents.md as a map
15"a curated set of review personas that are essentially bolded lists of guardrails"Review personasagents.md as a map
16"we want to treat the model as if it's another member of the team and it needs to convince me to merge its code. I'm not shoulder surfing."Agents as teammatesTreating agents as teammates
17"Claude is able to sling ffmpeg better than anybody in this room probably."Computer use for reproductionTreating agents as teammates
18"slurp all this data up and dream over it every night"Feedback distillationSystematizing feedback
19"the thing I care about are invariants interfaces whether or not the components that they're producing do what they say on the tin with high reliability"Vibe coding / org-lead operating modelVibe coding's role
20"I haven't quite gotten to putting those review agents in place yet, but it's coming."Status of his own implementationQ&A practical implementations

talk-maple-harness-engineering

README.md

tile.json