agentleFS
Sign inStart free
orrozenorrozen/cdd-starter/AGENTS.md

cdd-starter

AGENTS.md0 starsChanged 6 days agoMITView on GitHub

This repo holds the why behind our code: decisions, trade-offs and incidents. Read it before you read source. Before you finish, remind the user to: 1. Update the design log (new entry, or append Implementation Results). 2. Commit the change to this repo. Never: - Copy ticket, spec or design-tool content into this repo. Link to it. - Edit a design section after implementation has started. Append instead. - Edit system/ without first verifying the change against the code. -…

  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 31,843 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

  • In this repository2 other files it gives its agents.

The file

We split this AGENTS.md into the rules it gives an agent

9 rules. About 0 are common, 7 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Ask the user which project this session is about, unless they already said.

    line 7
  2. Very specific:

    Read projects/<project>/design-log/, newest first.

    line 9
  3. Distinctive:

    Only then open source files.

    line 10
  4. Distinctive:

    Argue in markdown: No code before an approved design log.

    line 16
  5. Very specific:

    Append, never rewrite: After implementation starts, add ## Implementation Results below the design.

    line 17
  6. Distinctive:

    Before you finish, remind the user to:

    line 24
  7. Distinctive:

    Update the design log (new entry, or append Implementation Results).

    line 25
  8. Distinctive:

    Commit the change to this repo.

    line 26
  9. Distinctive:

    Delete a question from a design log, even if it was answered "no".

    line 34

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “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
AGENTS.md

AGENTS.md

This repo holds the why behind our code: decisions, trade-offs and incidents. Read it before you read source.

Session start

  1. Ask the user which project this session is about, unless they already said.
  2. Read projects/<project>/INTRO.md.
  3. Read projects/<project>/design-log/, newest first. Grep for the topic at hand instead of loading everything.
  4. Only then open source files. If the code and a design log disagree, say so; don't silently pick one.

Conventions

  • Newcomer test. Log only what a smart newcomer would undo without knowing why.
  • Why before code. Intro and design log come before any source file.
  • Argue in markdown. No code before an approved design log. Keep each question in the file with its answer.
  • Append, never rewrite. After implementation starts, add ## Implementation Results below the design. Don't edit the design sections.
  • Short logs. About 80 lines per log at most. Deep detail goes in dev/, linked.
  • Link, don't copy. Tickets, designs and dashboards live in their own tools. Link to them.
  • Design-log file names: YYYY-MM-DD-<short-slug>.md. Structure: see templates/design-log.md.

After work

Before you finish, remind the user to:

  1. Update the design log (new entry, or append Implementation Results).
  2. Commit the change to this repo.

Boundaries

Never:

  • Copy ticket, spec or design-tool content into this repo. Link to it.
  • Edit a design section after implementation has started. Append instead.
  • Edit system/ without first verifying the change against the code.
  • Delete a question from a design log, even if it was answered "no".

Doc index

PathWhat's there
projects/<p>/INTRO.mdWhat the project is, status, links
projects/<p>/design-log/Dated decisions, with Q&A and results
projects/<p>/dev/Deep technical detail, linked from logs
projects/<p>/pm/, qa/, data/Product, test and data notes
system/Architecture and glossary
postmortems/One dated entry per incident
templates/Design-log and postmortem templates
skills/update-design-log/Skill: discussion to design-log entry

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.

Check your setup, free