AIQ Industries, LLC · Level 00 · The Grid
Ideas in.
Running systems out.
AIQ://OS carries every idea up one line, station by station, with a person at every signal. Here are the grid and the six stations in order, bottom to top. Ride one trip from the grid to the top.
The tower, level by level.
Level 00 · under every station
The Grid
AIQ://OS itself: rules, memory, routing and the harness.
- RulesHolds the rules, limits and review gates every agent is told to work inside.
- MemoryKeeps notes and decisions for good, and starts each session with only that workstream’s handoff.
- RoutingMatches each delegated job to a model and effort level, using a written routing table.
- UtilitiesConnects the tools AIQ Industries already runs on: Claude models, GitHub and Vercel hosting.
No signal at the grid
Ticket up: A new trip. Trip addresses are planned.
Read The Grid in fullLevel 01
Plan
Capture the idea and make it clear.
- CaptureLogs it with a date and a source, so nothing gets lost.
- ClarifyPulls in related notes and past decisions, so you never start from zero.
- StructurePlannedTurns it into a trip with its own address.
- Ready to exploreWrites a short handoff with only what Explore needs.
Demo signalGreen when it’s worth exploring
Ticket up: The goal and what done means.
Read Plan in fullLevel 02
Explore
Research the options before committing.
- QuestionsResearches open questions when asked.
- ResearchGathers sources and summarizes them, with links.
- ComparePlannedLays out the trade-offs. Running several AI models on the same brief, each unseen by the others, is planned.
- Select directionRecords the choice and the reason in the decision log.
Demo signalGreen when you pick a direction
Ticket up: The chosen direction and why.
Read Explore in fullLevel 03
Design
Decide how it should work.
- Map the systemDraws the parts and how they connect.
- Define connectionsWrites down what each part has to deliver.
- Trace the impactChecks what depends on what, and what could break.
- Approve the designRecords the approved plan. Later changes are logged as decisions.
Demo signalGreen when you approve the design
Ticket up: The approved design and what each part has to deliver.
Read Design in fullLevel 04
Build
Make the thing.
- Load the planLoads the approved plan and the context it needs.
- Make itAI agents write the code or the content.
- Test itRuns the tests and the build. Failures get fixed before review.
- Package itBundles it with notes on what changed and why.
Demo signalGreen when it works and you’ve seen it
Ticket up: The finished piece, what changed, and the test results.
Read Build in fullLevel 05
Review
Check it against what was asked.
- EvaluateA separate reviewer, not the builder, checks the work.
- Resolve findingsFixes what you agreed to, then checks again.
- Check the evidenceAttaches test results, previews and sources.
- Release decisionShips only on a yes. A no goes back to Build with the reason.
Demo signalGreen only on your yes
Ticket up: The release decision and the fixes that were made.
Read Review in fullLevel 06
Maintain
Keep it healthy, learn from it, go again.
- ObserveChecks the site’s copy for claims that have gone stale, on request.
- Find patternsLogs repeated mistakes, so the fix goes into the process.
- Reuse what worksPackages it so every future trip benefits.
- Frame the next tripOpens a new Plan with what was learned attached.
Demo signalGreen when you choose what’s next
Ticket up: What was learned, attached to the next Plan.
Read Maintain in full
This signal is a demo. On a real trip it turns green only when a person says go. Green is reserved for exactly that.
Now · updated
Building AIQ://OS, this site, and getting the business running.
- AIQ://OSIN USE
- KYD BaseIN DESIGN
- ConsultationsOPEN
Latest from the work log:Ran the launch checks on the new site
Knowledge · the blog
What we learn, we publish.
Banks Are Trying to Wall Off Third-Party AI Agents. The Hard Part Is Proving Who Asked for the Transaction.
Morgan Stanley and Bank of America are reportedly trying to contain unvetted AI agents touching customer accounts, while Anthropic ships a proactive agent app and METR and Redwood find holes in OpenAI's agent monitors. Attribution is the unsolved piece.
One Operator's AI Agents Hit 1,200 Sites and Breached 100+ Companies. The Cost Was About $50 Per Success.
Researchers at NorForge found exposed control panels for Cairn, an open-source pentesting agent tool, being used to attack shops and sites across four countries. The success rate was only ~10%. The economics are the story.
After a 34% Drawdown, BlackRock Is Filing a Bitcoin ETF That Sells Downside Protection
BlackRock's third bitcoin ETF filing promises upside exposure with a level of downside protection. The mechanism, cap and fee aren't public yet, and the Swiss spot-ETF approval rests on one outlet. What's solid, what isn't, and what we're watching.
Work with us
Building something, or trying to understand it?
Either way, this is where it starts: a free first conversation to find out what is actually worth building. AIQ Industries can set AIQ://OS up inside your business, answer plainly where AI fits, or build a custom system you own outright.