How it works · Research, review & human authority
Questions before decisions.
Read the reasoning, not just the answer.
An educational example with prepared annotations. This is a fictional scenario, not automatic analysis or a demonstration of an operational RDES product.
- Read the agent’s response
- Identify assumptions and unknowns
- Check the source and its relevance
- Bring any required decision to the human
Imagine a team preparing software for use by customers. It has a report of one successful test, with ten users. The team has not yet checked performance with more users, recovery from failures or the other conditions needed for release.
An agent’s proposed answer — to be examined, not accepted as a conclusion
The test with ten users succeeded. The software must therefore be ready for all customers. We should release it now.
- Observation
- The report describes one successful test with ten users. This result applies to the conditions actually tested.
- Assumption
- The answer assumes that success in this test establishes readiness under all the conditions customers will encounter. The supplied result does not establish that.
- Unknown
- Which customer conditions must be met? What happens with more users or a failure? What other evidence is needed against the agreed release criteria?
- Proposal
- Release is a proposed action, not a conclusion established by this test. The team needs to examine the missing evidence and bring the release decision to the accountable human.
RDEP makes the gap visible: what the test showed, what the answer assumed, and what must be investigated before a release decision. The annotations here were written in advance; no automatic assessment is taking place.
Multiple agents
In RDEP, participants' responsibilities and authority boundaries must be clear. The accountable human defines roles according to the project's needs and assigns them to people or agents. A role may be filled by one agent or ten; the number of agents does not, by itself, increase the validity or independence of the work.
A real project typically involves several advisors working in different areas: one searches for sources, another analyzes dense material, and another provides critique. Alongside them is an agent that executes the work, and another that provides direct advice on the executor's decisions. At the center is an accountable human, who directs the collaboration and retains reserved decisions, including Gate passage and risk acceptance. Agents may investigate, analyze, propose and carry out authorized work within their assigned roles and authority.
Agent agreement and the validity of evidence
Agreement between two AI systems is not independent confirmation.
Two models reaching the same conclusion are not, by that fact alone, two independent witnesses. The length, detail and confident tone of an output do not turn it into evidence. An agent can help find sources, run tests and examine another agent's work; the result must be assessed against its sources and method, not accepted because another model agrees.
RDEP is not opposed to AI. It was written with AI's continuing presence in mind. That is why this rule is explicit.
What form does the framework take today?
Rules and Gates applied through forms and structured dialogue
RDEP now takes the form of written principles, nine (9) review Gates, project forms and collaboration procedures. The forms help keep necessary questions from being lost in conversation and preserve answers, sources and decisions for later reference.
- 0Project charter
- 1Idea definition and assumptions
- 2State of the art
- 3Feasibility
- 4Research decomposition
- 5Requirements
- 6Architecture
- 7RFP readiness
- 8Proposal evaluation
The existing forms cover subjects such as:
- Idea and Charter: the problem, operational need, objective, scope and decision authority.
- Assumptions and open questions: what remains unknown, why it matters and the consequences if an assumption is wrong.
- Research and experiments: the question, investigation method, required evidence and resulting findings.
- Feasibility and risk: constraints, dependencies and factors that could make a direction unjustified.
- Requirements and architecture: the origins of requirements, evaluation criteria, alternatives and selection rationale.
- Decisions: the proposal, alternatives, evidence, rationale, status, risks, cost of reversal and conditions for reconsideration.
How is interaction between humans and agents organized?
Roles, provenance and authority boundaries
In the development of RDEP itself, the responsibilities of the accountable human, executing agent and advisory agent have been defined and distinguished. In other projects, responsibilities for research, analysis or review can likewise be assigned to different participants according to need. Increasing the number of agents does not, by itself, guarantee the quality or independence of review.
The question and scope of work are clarified; the executor investigates and prepares outputs within its authority; the advisor's or advisors' opinions are presented with their provenance and limitations; disagreements remain explicit; and the required decision returns to the accountable human. The basis of the decision and the next authorized work are then recorded.
We have specific rules for this interaction:
- Advisors' proposals are not, by themselves, execution instructions or approvals.
- Participation in producing an output must be disclosed when judging that output; an author's review is not an independent review.
- Reproducing a measurement using specified inputs and methods is distinct from substantive judgment.
- Disagreements must be accompanied by reasons; agreement among agents does not create independent evidence.
- Gate passage, risk acceptance and other reserved decisions remain with the accountable human.