Skip to content
production / main000%
All tools
ADR-01DOC GENERATOREVIDENCE-FIRST / NO LOGIN

Architecture Decision Record Generator

Fill the form and generate a concise Markdown Architecture Decision Record ready for a team or coding agent.

01Decision context02Alternatives03Reversal condition
ADR-01 / EVIDENCE WORKBENCH
OBSERVED → DERIVED → DECISION

A report can be previewed, but without a target name it is not a shareable technical record.

ADR / MARKDOWN
# ADR: Architecture decision

**Status:** Accepted
**Date:** 2026-08-10

## Context
—

## Decision drivers
—

## Options considered
- —

## Decision
—

## Consequences and trade-offs
—

## Reversal / revisit condition
—

## Evidence
—

## Assumptions
—

## Validation plan
—
AC / EVIDENCE REPORT
AC-ADR-01-20260810100627 · METHOD 2.0

PDF / human · MD / repo · JSON / automation

ANALYSIS SUBJECT
SELF-REPORTED
TARGET NOT PROVIDED
TARGET
Architecture decision
SCOPE
Architecture Decision Record
EVIDENCE BASIS
SELF-REPORTED
ADR-01 / EVIDENCE FILE

Architecture Decision Record Generator

5 decision-evidence gaps

EVIDENCE COVERAGE
45/100
CONFIDENCE
LOW
CLAIM BOUNDARY

An ADR does not declare a decision correct; it makes context, alternatives, evidence, consequences and revisit conditions traceable.

01

Executive brief

Decision evidence and validation plan are first-class ADR fields.
Revisit conditions prevent accidental permanent lock-in.
02

Findings and decision points

F-1HIGH CONFIDENCE

Context missing

The decision can be recorded but should not be treated as strong until the evidence chain is complete.

EVIDENCE → E-CTX · E-EVIDENCE · E-VALIDATION
F-2HIGH CONFIDENCE

Drivers missing

The decision can be recorded but should not be treated as strong until the evidence chain is complete.

EVIDENCE → E-CTX · E-EVIDENCE · E-VALIDATION
F-3HIGH CONFIDENCE

Options missing

The decision can be recorded but should not be treated as strong until the evidence chain is complete.

EVIDENCE → E-CTX · E-EVIDENCE · E-VALIDATION
F-4HIGH CONFIDENCE

Evidence missing

The decision can be recorded but should not be treated as strong until the evidence chain is complete.

EVIDENCE → E-CTX · E-EVIDENCE · E-VALIDATION
F-5HIGH CONFIDENCE

Validation plan missing

The decision can be recorded but should not be treated as strong until the evidence chain is complete.

EVIDENCE → E-CTX · E-EVIDENCE · E-VALIDATION
03

Evidence ledger

E-CTXUSER INPUT
Decision context
CONF LOW
E-EVIDENCEUSER INPUT
Decision evidence
CONF LOW
E-VALIDATIONUSER INPUT
Validation plan
CONF LOW
04

Method, assumptions and boundaries

05

Recommended next move

A01Close evidence gaps and run the validation plan.
A02Reopen the ADR when its revisit condition is met.
This report is decision support. Observations, derivations, user inputs and heuristics are deliberately separated.
AC-METHODEVIDENCE → DECISION

The report is not sales decoration; it is a technical decision record.

Each report separates observations, user input, derived conclusions and heuristic assumptions. If evidence is weak, confidence drops.

WHEN TO USE IT
  • Prevent architectural decisions from losing their rationale after a few weeks
  • Give a team or coding agent a short, durable architecture boundary
  • Record alternatives and explicit conditions for revisiting the decision
METHOD / INTERPRETATION

Why should an ADR stay short?

An ADR is not a full design document. Its job is to preserve the context and trade-offs of one decision. Short records are more likely to stay current; long documents are often abandoned when the decision changes.

Why record a revisit condition?

Good architecture decisions do not need to stay correct forever. Recording the traffic, team, cost or product signal that would trigger a revisit turns tool-switching into an explicit condition rather than an impulse.

BOUNDARY

Generated Markdown is not the decision itself; the actual alternatives and trade-offs still need human validation.

If this output needs to become a real technical decision, move it into the FORGE flow.

Open FORGE

AVAILABLE FOR PROJECTS · PRODUCT & SYSTEMS ARCHITECT · TAKEOVER / STABILIZE / OPERATE · WEB · ANDROID · WINDOWS · AHMET CANAL

Contact

Has your product grown faster than its system?

Send the current situation, your biggest blocker and the outcome you want. We will clarify scope together.

Availability

Open to new consulting and project-based work.

GROWTH / SIGNAL LOOP

Move from reading to a decision.

Booking, feedback and system notes in one low-friction exit.

CNV-01

Technical call

If there is a fit, move directly to a 20-minute call.

CNV-02

Decision blocker

Tell me what is still missing with one signal.