mcpbeat Sign in

Adr Drafting Agent Skill

Creates new Architecture Decision Record (ADR) documents for significant architectural changes using a consistent template and repository-aware naming and storage guidance. Use when a user or agent decides on an architectural change, needs to document technical rationale, or wants to add a new ADR to the project history.

2k tokens
context cost
the whole folder, loaded on every use
3
files
instructions only
0
copies elsewhere
how many repositories repackaged it
316
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository
npx skills add https://github.com/giuseppe-trisciuoglio/developer-kit --skill adr-drafting

The instruction itself

14 sections, as written by the author

ADR Drafting

Creates new Architecture Decision Record (ADR) documents for major architectural choices so teams can keep a clear history of why important technical decisions were made.

Overview

This skill helps create a new ADR from discovery to final markdown file. It confirms the decision details, inspects the repository for any existing ADR conventions, and drafts a new ADR with the standard sections Title, Status, Context, Decision, and Consequences.

When the repository does not already have an ADR convention, default to storing ADRs in docs/architecture/adr and use a zero-padded filename such as 0001-use-postgresql-for-primary-database.md.

See references/template.md for the default ADR template and references/examples.md for example ADRs and naming patterns.

When to Use

Use this skill when a user or agent has decided on a meaningful architectural change and needs to document the rationale, chosen direction, and trade-offs in a new Architecture Decision Record. It fits requests such as creating an ADR, documenting an architecture decision, writing a decision record, or preserving the project history behind an important technical choice.

Instructions

Phase 1: Confirm the ADR inputs

Ask the user for the minimum information needed to draft a new ADR:

  • Decision title
  • Decision status (Proposed by default if not yet finalized)
  • Context: the problem, constraints, or forces driving the decision
  • Decision: the chosen approach
  • Consequences: what becomes easier, harder, riskier, or more expensive
  • Confirm that this request is for a new ADR, not for editing an existing ADR
  • Confirm the desired repository language if documentation language is unclear
  • Confirm whether any existing ADR naming convention must be preserved

If the user actually wants to update an existing ADR, change statuses in older ADRs, or manage supersession links, explain that this skill only drafts new ADR documents and ask whether they want to proceed with a new record instead.

Phase 2: Discover ADR conventions in the repository

Inspect the repository before drafting:

  • Search for likely ADR locations such as:
  • docs/architecture/adr
  • docs/adr
  • adr
  • architecture/adr
  • If ADR files already exist, read one to three examples to infer:
  • numbering format
  • filename pattern
  • title format
  • language and tone
  • If no ADR directory exists, recommend docs/architecture/adr
  • Determine the next ADR number from existing files when possible
  • If no prior ADR exists, start with 0001

Preferred default naming when no convention exists:

  • Directory: docs/architecture/adr
  • Filename: NNNN-short-kebab-title.md
  • Title: # ADR-NNNN: <Decision Title>

Phase 3: Draft the ADR

Create a draft using the standard structure:

# ADR-NNNN: Decision Title

## Status
Proposed

## Context
What problem, constraints, or trade-offs led to this decision?

## Decision
What architectural choice was made?

## Consequences
What becomes easier, harder, riskier, or more expensive because of this decision?

Drafting rules:

  • Keep the title specific and decision-oriented
  • Capture enough context to explain *why* the decision was needed
  • Record the chosen direction clearly and directly
  • Include both positive and negative consequences when known
  • Do not invent rationale, constraints, or outcomes that the user did not provide
  • If critical information is missing, insert concise placeholders or ask follow-up questions before finalizing

Phase 4: Review the draft with the user

Before writing files, present:

  • proposed file path
  • proposed title
  • ADR status
  • a concise preview of the drafted sections

Ask for approval before creating the file. If the user wants adjustments, revise the draft first.

Phase 5: Create the ADR file

After approval:

  • Create the ADR directory if it does not exist
  • Write the ADR markdown file using the repository's established pattern when available
  • Preserve the user's wording for decision rationale as much as possible while keeping the document concise
  • Report the final file path and summarize what was created
  • Stop after creating the new ADR file so the skill remains focused on a single new decision record

Examples

Example 1: New database decision

User request: "Create an ADR for moving from SQLite to PostgreSQL"

Expected flow:

  • Confirm the title, status, reasons for the change, and expected consequences
  • Check whether the repository already has ADR files
  • Draft a new ADR in the existing convention or default to docs/architecture/adr/0001-move-to-postgresql.md
  • Ask for approval before writing the file

Example 2: New service boundary

User request: "Document the decision to split billing into a dedicated service"

Expected flow:

  • Ask for the architectural context and why the current design is insufficient
  • Confirm the chosen boundary and the operational consequences
  • Draft a new ADR with the standard sections
  • Create the file only after user approval

See references/examples.md for longer ADR examples.

Best Practices

  • Keep each ADR focused on one architectural decision
  • Match existing repository naming, numbering, and writing style when ADRs already exist
  • Prefer concise explanations over long narratives
  • Capture trade-offs honestly, including downsides and new risks
  • Default the status to Proposed unless the user confirms another state
  • Use the repository's preferred documentation language when it is clear

Constraints and Warnings

  • This skill is for new ADR creation only
  • Do not update existing ADR files, status histories, or supersession chains
  • Do not fabricate missing rationale or consequences
  • Do not force docs/architecture/adr if the repository already uses another ADR location
  • Ask clarifying questions whenever the decision, context, or consequences are too vague to document responsibly

Other skills for the same job

different authors, same section of the catalogue
Markdown To HTML
by github
vendor ×1

Convert Markdown files to HTML similar to `marked.js`, `pandoc`, `gomarkdown/markdown`, or similar tools; or writing custom script to convert markdown to html and/or working on web template systems like `jekyll/jekyll`, `gohugoio/hugo`, or similar web templating systems that utilize markdown documents, converting them to html. Use when asked to "convert markdown to html", "transform md to html", "render markdown", "generate html from markdown", or when working with .md files and/or web a templating system that converts markdown to HTML output. Supports CLI and Node.js workflows with GFM, CommonMark, and standard Markdown flavors.

29k tokens
Documentation
by anthropics
vendor ×1

Write and maintain technical documentation. Trigger with "write docs for", "document this", "create a README", "write a runbook", "onboarding guide", or when the user needs help with any form of technical writing — API docs, architecture docs, or operational runbooks.

376 tokens
Visa Doc Translate
by mturac
×1

Translate visa application documents (images) to English and create a bilingual PDF with original and translation

1k tokens
Docs Writer
by christophacham
×1

Write, review, and edit documentation files with consistent structure, tone, and technical accuracy. Use when creating docs, reviewing markdown files, writing READMEs, updating `/docs` directories, or when user says "write documentation", "review this doc", "improve this README", "create a guide", or "edit markdown". Do NOT use for code comments, inline JSDoc, or API reference generation.

2k tokens
Constitution Guardian
by ComeOnOliver
×1

Real-time Constitution compliance checker for devflow documents. Blocks partial implementations and hardcoded secrets during file editing.

5k tokens
Global Commenting
by ComeOnOliver
×1

Write self-documenting code with minimal, evergreen comments that explain complex logic without describing recent changes or temporary fixes. Use this skill when writing code comments, documentation strings, explaining complex algorithms, clarifying business logic, or deciding whether code needs comments. Apply when working with any source code files where comments or documentation might be added, ensuring comments remain relevant, helpful, and focused on explaining why rather than what the code does, while preferring clear code structure and naming over excessive commenting.

3k tokens
Hads
by wshobson

Use when writing technical documentation that needs to be readable by both humans and AI models, converting existing docs to HADS format, validating a HADS document, or optimizing documentation for token-efficient AI consumption.

2k tokens
Ic Memo
by anthropics
vendor

Draft a structured investment committee memo for PE deal approval. Synthesizes due diligence findings, financial analysis, and deal terms into a professional IC-ready document. Use when preparing for investment committee, writing up a deal, or creating a formal recommendation. Triggers on "write IC memo", "investment committee memo", "deal write-up", "prepare IC materials", or "recommendation memo".

743 tokens

How to use it

Copy the folder

Take giuseppe-trisciuoglio/adr-drafting from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.