Working with Robots
Your engineers picked up AI team by team, ahead of any shared plan. Working with Robots is our framework for leading that work: four domains, twelve disciplines, and the investment tiers for deciding where your energy goes next.
The framework starts with a call only you can make.
First, decide what to hand over.
There is one decision the framework cannot make for you: which kinds of work belong to a robot at all. That is a leadership call, not a setting you flip on. Handing an agent your test coverage is a different bet than handing it a customer-facing feature.
Some of that work is worth handing over today and some is not, and the line moves: last quarter's too-risky is this quarter's obvious yes. Decide where robots belong before you invest a thing in making that work go well.
Then, make the work go well.
An agent does useful work when three things are true at once: it has what it needs to act (Context), it can only do what it should (Control), and you can see what it did afterward (Observability). The fourth domain, Improvement, works differently: it is how your team gets better over time.
The first three are conditions, not steps: where one is weak, the work suffers for it. Each domain breaks into three disciplines, the named units of practice a team can point at and fund.
The four domains.
The moment your engineers started using AI, your org took a position in all four, whether anyone chose it or not. Read each one to find the position you actually hold, and what the next investment buys.
Context
Faster delivery with less rework, from agents that start with the right information.
Control
Moving fast without betting the company on an agent's judgment.
Observability
Trust backed by evidence: reconstruct any incident, answer any auditor from the record.
Improvement
An AI investment that compounds instead of plateauing.
Get your hands on it.
You don't have to read the whole framework to use it. Place your bets on a hypothetical team and read the consequences, or download the whole thing and hand it to your own agents.
Feel it in play.
The game hands you a hypothetical: your team is shipping an authentication service with AI in the loop, and you have eight chips of investment to spread across the twelve disciplines. Place them and you get back a memo reporting how the next six months went under those bets. Move a chip and the memo rewrites, so you can feel exactly where the trade-offs live.
Take it with you.
The bundle is the whole framework in files your agent can read. Give it to Claude, Cursor, or whatever you already use, then ask where your own org stands. It is plain reference material: no code, nothing that runs.
Or install it with one line
mkdir -p ~/.claude/skills && curl -fL https://gobetweenlab.com/working-with-robots/working-with-robots-framework.tar.gz | tar -xz -C ~/.claude/skillsSee it at work.
The framework maps where an org invests; the same disciplines recur around a single AI system. A field guide follows one practice across the map, from nothing to steady state.
The investment tiers.
Every discipline sits at one of four investment tiers. The tiers are not rungs to the top: each buys you something and costs more to run and to change than the last, so the tier you want is the one that fits your org, not the deepest.
So for each of the twelve disciplines, the question is the same: which tier are we at, and is the next one worth buying?
Rather feel it than read it? Place your bets.
Here is how it all works on one discipline.
Collection
An agent can only act on what it can see. Collection is the discipline of putting the right material in front of it: the meeting transcripts where a decision actually got made, the code and its git history, the wiki or knowledge base, the epics and tickets that say what the work is for.
Select a tier
People grab whatever artifacts they think the agent needs, with no shared standard. It is where every team starts.
- Enables
- The cheapest look at what your agents actually need. Every session someone hand-feeds shows which artifacts an agent reached for and which it ignored, the knowledge a durable setup gets built from.
- Deficiencies
- Critical artifacts get missed and stale ones get pulled in, and no one can say what is actually available.
- Cost to maintain
- The quiet tax of re-gathering the same context, session after session.
Techniques
None yet. Manual and ad hoc.
Five principles for working with robots.
By the time AI reaches your desk, it looks like a tools question: which agent to standardize on, which vendor to believe, what the demos are worth. The decisions that will still matter next year are judgment calls: where to invest, and what each choice costs. These five principles are that judgment written down. The whole framework, domains to disciplines to tiers, is built on them.
Humans make decisions
Robots contribute. Humans decide.
Sometimes the robot executes, sometimes it thinks alongside you, but humans hold judgment where it matters: approving plans, reviewing output, deciding what ships.
Diverge then converge
Close the loop before moving on.
Agents follow patterns imperfectly and accumulate entropy. After a round of divergent work, realign to your patterns; skip convergence and the mess compounds.
Fluency isn't value
The sensation of work is not the work.
AI manufactures the sensation of progress, convincingly enough to stop you checking whether the output is useful. Measure against a standard the system can't generate.
Input is the lever
You don't own the output. You own everything around it.
What you control is the intent and information you feed in, the tools the agent can reach, and the checks its work has to clear, all set before the run starts. Build those into the system and they pay off on every run; fix each output by hand and you are only doing cleanup.
Map, not a scorecard
A map tells you where you are, not how you rank.
The four domains and twelve disciplines are a useful fiction: the real work overlaps, and investing on both sides of an overlap beats either alone. The tiers are real but not grades: each costs more than the last to run and to change, so the right tier is the one that fits your org, not the deepest one. And the map is of your terrain, not an ideal to conform to, so the right shape is something you discover by working up from the bottom, not a phase you skip to the top.
Bring this to your org.
The framework is how we think. Deciding where your team invests next, and leading them through the trade-offs, is what we do. Our diagnostic locates your org on this map, tier by tier: see where you stand, then plan the next quarter with us.