agentleFS
jakubsuplickijakubsuplicki/codument/skills/tdd/SKILL.md

tdd

Skill48 starsChanged 40 days agoApache-2.0View on GitHub

Implement a planned Codument work step with a feedback-first, red-green-refactor loop where practical.

195 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use capture-learning-tools, design-frontend-dev and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • The-Vibe-CompanySkill

    capture-learning-tools

    Turn technical conversations, bugs, CI failures, review comments,

    Also in 5 projects that use this

  • The-Vibe-CompanySkill

    design-frontend-dev

    Clean Impeccable frontend design skill. Use when the user wants to

    Also in 5 projects that use this

  • tristanphvnSkill

    ketchup-video

    Catch the user up like /keel:ketchup, with a short silent video of slides for what happened and the waiting calls as text under it. Use for /keel:ketchup-video.

    Also in 14 projects that use this

  • jaydubya818Skill

    figure-it-out

    Use when novel or multi-phase work needs an auditable workflow and no narrower playbook fits.

    Also in 54 projects that use this

  • jaydubya818Skill

    principle-laziness-protocol

    Apply when a refactor is adding wrappers or abstractions instead of simplifying.

    Also in 45 projects that use this

  • +2 more from The-Vibe-Company
  • +5 more from jaydubya818

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 187 edited versions we read keep these.

  • 7

    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 this when implementing a planned step or fixing a bug.

    line 8
  2. Distinctive:

    Read the approved plan step and its mapped docs.

    line 12
  3. Distinctive:

    Update docs and registry when behavior or source ownership changes.

    line 25
  4. Distinctive:

    Do not write all tests first and then all implementation.

    line 52
  5. Distinctive:

    Do not refactor while red.

    line 53
  6. Distinctive:

    Do not add speculative behavior for later steps.

    line 54
  7. Distinctive:

    If no correct test seam exists, document that finding and use the best available feedback loop.

    line 55

“Common” is measured: kept by at least 6 in 10 of the 187 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

Test-Driven Development

Use this when implementing a planned step or fixing a bug. The principle is feedback first: build the fastest reliable loop that proves the behavior.

Workflow

  1. Read the approved plan step and its mapped docs.
  2. Identify the plan's promised milestone and the user or integration boundary that supplies its acceptance evidence. Exercise a representative end-to-end experience early; supporting infrastructure names the milestone it unblocks.
  3. Choose the cheapest reliable feedback loop that exercises that boundary; use a small representative experiment before building expensive supporting infrastructure:
    • Unit test for pure logic
    • Integration test for module boundaries
    • CLI invocation for command behavior
    • Browser or UI harness for interface behavior
    • Repro script for a bug when a proper test seam does not exist yet
  4. Prefer red-green-refactor:
    • Red: write one failing behavior test
    • Green: add the smallest code change that passes it
    • Refactor: improve structure only while tests are green
  5. Repeat one behavior at a time.
  6. Update docs and registry when behavior or source ownership changes.

Evidence boundary

Record what was exercised, what was observed, and what remains untested. A mock, fake, or local executor supports only its exercised contract; it does not prove an external provider's behavior, downstream checks, or an end-to-end effect. Use the real integration when the approved acceptance criteria require it and it is available within the authorized scope. If it is unavailable, keep that criterion open, record the blocker, and bring any change to acceptance back to the user. A green local suite cannot silently replace required integration evidence.

An expected failure during the red phase is feedback, not a final verification failure. Once the relevant checks pass, broaden them only for required checks or a concrete unresolved concern; do not build a new harness for a reversible edit that an existing check can verify.

Routine test, mapping and documentation failures call for diagnosis, correction and re-verification within the approved contract. Keep required acceptance open while evidence is missing. Ask the human when repair requires an unapproved decision, unavailable external input or genuinely irreversible action; preserve explicit pauses, gated mode and no-commit requests. A new file in an outcome plan needs routing and ownership, not renewed scope approval; legacy file-bound approval remains strict.

Test Quality

Good tests verify behavior through public interfaces. They should survive internal refactors.

Avoid tests that:

  • Assert private implementation details
  • Mock internal collaborators when a real boundary is cheap enough
  • Encode imagined future behavior before the current slice teaches you what matters
  • Pass while the user-visible behavior is broken

Rules

  • Do not write all tests first and then all implementation.
  • Do not refactor while red.
  • Do not add speculative behavior for later steps.
  • If no correct test seam exists, document that finding and use the best available feedback loop.

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.