Principles for governing AI-assisted research.

The problem

Researchers increasingly share AI workflows as recipes: sequences of prompts, steps and tools. A recipe rarely says how a team should check its outputs, or who may accept them. Generating candidates can be cheaper and faster than evaluating them, and finished-looking output can conceal substantive defects. The paper behind TRACE-OS argues that AI-assisted research needs principles for organizing its workflows more than it needs further recipes.

The five governing principles

The principles come from software engineering and security research, where teams of people and machines have long integrated separate pieces of work. They govern the relevant actions whether a person or an agent performs them, and they rest on one assumption: a researcher remains responsible for the outcome. For each principle, TRACE-OS keeps the record listed beneath it.

1. T

   Track records

   Record what happened, in what order, and why, and keep the work’s current state visible to everyone who uses it.

   Shared version history, and a trace record for every agent session that commits work, linking the captured instructions to the agent’s choices, with their reasons, and to the changes they produced.

2. R

   Review against sources

   Compare an output with an explicit criterion or another source before others rely on it. Sources include the data and the logic of a claim, not only citations.

   Checking records that name the candidate, criterion, sources, method, outcome and limits; model objections must quote the text, and the quotes are checked.

3. A

   Accept independently

   Give each unit of work an owner, and separate producing the work from the decision to accept, return or refuse it, taken by an actor other than its producer. A researcher who authorizes an agent or a process to accept stays accountable for what it accepts.

   Decisions by a person, or by an agent or process under a named researcher’s authority, recorded against the version they cover and apart from model critique.

4. C

   Control transitions

   Check and authorize outside material and AI output before they enter the manuscript, the project’s data or its standing instructions.

   Intake records of who admitted, returned or refused outside material, which waits apart until then, and transition records for drafts moved into the shared manuscript and instructions adopted as defaults.

5. E

   Evaluate corrections

   Take what review or use finds back into the work, and keep a detected error open until the team has checked that the correction worked.

   A defect ledger linking findings, claimed fixes, recurrence and targeted rechecks of the original problem, one for each output a finding reaches.

How the workflow runs

1. 1Direct the taskIn your own words, kept verbatim.T
2. 2The agent executesAnd drafts a record of its choices and checks.T
3. 3Check the candidateAgainst its sources.R
4. 4DecideAn independent actor accepts, returns or refuses it.A

Nothing is blocked: work can be committed and shared before anyone decides on it, and it remains visibly unaccepted until someone does. The workflow describes each step and the boundary transitions between them.

What the records establish

The records show which actions were taken on a unit of work and what evidence was retained. They make gaps visible, because a missing action is reported rather than refused. They do not establish that every required action occurred, that a check was adequate, or that a decision was sound. Whether the workflow improves research outcomes remains an empirical question.

Where to go next

Get startedSet up a project and record your first unit of work.

The workflowHow a unit of work moves from direction to decision, and what happens at each boundary.

The five principlesWhat each principle requires, and what TRACE-OS keeps for it.

Where TRACE comes fromThe paper behind the principles, and how they were derived.