type-test-code
Skill20 starsChanged 43 days agoMITView on GitHub
Write types first, then test and code independently
In this repository26 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- breakdownSkill
- bug-investigationSkill
- code-review-adversarialSkill
- code-review-multiSkill
- code-review-with-followupSkill
- document-changesSkill
- grillingSkill
- explain-changesSkill
- github-actions-ciSkill
- github-appSkill
- github-ticketsSkill
- implementation-planSkill
- implement-simpleSkill
- implementSkill
- implement-ttcSkill
- orchestrationSkill
- pull-requestSkill
- review-and-wrap-upSkill
- verusSkill
- specSkill
- ticketsSkill
- update-working-copySkill
- verifySkill
- wrap-upSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
15 rules. About 0 are common, 5 are distinctive, and 10 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
10
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 8
write tests with an indepdent subagent: done by the tester.
- Very specific: line 9
write code independent of the tests: done by the coder.
- Very specific: line 15
Use the /orchestration skill to manage sub-agents.
- Very specific: line 20
If you have already been given a plan that specifies these changes, go ahead and use it.
- Very specific: line 21
If instead you need to determine the changes to make, use the planner sub-agent to implement the types.
- Distinctive: line 27
Ensure the code type checks.
- Very specific: line 32
If there is no tester sub-agent role defined than use the implementer sub-agent role.
- Distinctive: line 34
Write the tests without looking at the implementation, just the types.
- Very specific: line 36
Use /verify to guide the writing of tests.
- Distinctive: line 38
If tests are to be written in the same file as the code, see if you can first write tests in a separate file and then later move them into …
- Very specific: line 41
When notified that an implementation section has landed, run the relevant tests.
- Very specific: line 42
When there is a test failure due to an implementation issue, notify the coder.
- Distinctive: line 50
Write the implementation for the types without looking at the tests.
- Very specific: line 52
When a section of code writing is completed, notify the tester sub-agent.
- Distinctive: line 53
When you are notified about a test failure, you can then look at the test and run it yourself.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill 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.
Files
When writing code
- first design and implement the types: done by the
planner. - write tests with an indepdent subagent: done by the
tester. - write code independent of the tests: done by the
coder.
The task is complete when both
- the
coderhas completed the implementation plan (written all the code) - the
testeris satisfied with the level of testing.
Use the /orchestration skill to manage sub-agents.
Planner
Make all type changes at testing seams.
If you have already been given a plan that specifies these changes, go ahead and use it.
If instead you need to determine the changes to make, use the planner sub-agent to implement the types.
This includes either interfaces or function/method stubs. You may need to write stub code inside a function to satisfy type checking: use either a standard for "unimplemented" or a TODO comment so the incompletion is obvious.
Ensure the code type checks.
Tester
Once types are in place, spawn a separate tester sub-agent for writing tests.
If there is no tester sub-agent role defined than use the implementer sub-agent role.
Write the tests without looking at the implementation, just the types.
Use /verify to guide the writing of tests.
If tests are to be written in the same file as the code, see if you can first write tests in a separate file and then later move them into the code file after the code implementation is complete.
Maintain a persistent testing sub-agent to continue to run the tests until they all pass.
When notified that an implementation section has landed, run the relevant tests.
When there is a test failure due to an implementation issue, notify the coder.
Coder
Spawn a separate coder sub-agent for writing the code.
If there is no coder sub-agent role defined than use the implementer sub-agent role.
The sub-agent should be given an /implementation-plan for the changes.
Write the implementation for the types without looking at the tests.
When a section of code writing is completed, notify the tester sub-agent.
When you are notified about a test failure, you can then look at the test and run it yourself.
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
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.