agentleFS
daniel-kindldaniel-kindl/dkstack/skills/write-adr/SKILL.md

write-adr

Skill1 starsChanged 9 days agoMITView on GitHub

Use when you make a decision that is durable or hard to reverse, such as a framework, data store or public contract. Creates an Architecture Decision Record.

12 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 12 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use when a design decision is durable or hard to reverse.

    line 8
  2. Distinctive:

    Create an Architecture Decision Record using architecture/adrs.md.

    line 10
  3. Distinctive:

    Write a short ADR with:

    line 15
  4. Distinctive:

    Do not rewrite old ADRs. Add a new one if the decision changes.

    line 21
  5. Distinctive:

    If you can start subagents, use them to research the options.

    line 25
  6. Distinctive:

    If you cannot start subagents, do the steps one at a time in the order of this skill.

    line 36

“Common” is measured: kept by at least 6 in 10 of the 12 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Write ADR

Use when a design decision is durable or hard to reverse.

Create an Architecture Decision Record using architecture/adrs.md.

Steps

  1. Confirm the decision is worth recording (affects multiple components or is costly to change).
  2. Write a short ADR with:
    • Context
    • Options considered
    • Decision
    • Consequences
  3. Store it with the code it affects.
  4. Do not rewrite old ADRs. Add a new one if the decision changes.

Subagents

If you can start subagents, use them to research the options. Do step 1 yourself. Then research each option for the "Options considered" section at the same time, each option in its own subagent. Give each subagent a brief that is complete without your context:

  • The decision and its context.
  • The one option that it researches, and the names of the other options.
  • The constraints that you know, for example the components that the decision affects.
  • What to report: what the option does, its advantages, its disadvantages and its cost to reverse. Each claim has a source, for example a file path or a link.
  • A word limit, for example 300 words.
  • This line: "Report to me only. Do not write the ADR, post, approve or merge."

Then collect the reports. Write the ADR yourself in step 2, and do steps 3 and 4 yourself. A report from a subagent is a claim. Check each claim that the decision depends on at its source. If a claim has no source, do not put it in the ADR.

If you cannot start subagents, do the steps one at a time in the order of this skill. Research one option fully before you start the next option.

Rules

  • Keep it short.
  • State trade-offs.
  • Link related principles from architecture/ when useful.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.