azclips / Agent
AGENTS.md15 starsChanged 2 days agoMITView on GitHub
Scoped guidance for the AI agent under Agent/ and its tests. Read this before changing Agent code or Agent tests. When a rule here changes, update this guide in the same PR. - Build expected paths from the product's path constants (for example HomePaths), not from string literals. A literal path can drift from what the product writes, and the assertion then passes for the wrong reason. - Prove that a new assertion fails when the condition it guards is…
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 31,439 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 repository13 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .azure-pipelines/AGENTS.mdAGENTS.md
- devops/scripts/AGENTS.mdAGENTS.md
- azclips-custom-code-classifierSkill
- azclips-scenario-test-converterSkill
- extension-test-coverageSkill
- migrate-clips-to-extensions-codegen-llm-hybridSkill
- openspec-apply-changeSkill
- openspec-archive-changeSkill
- openspec-exploreSkill
- openspec-proposeSkill
- openspec-sync-specsSkill
- openspec-verify-changeSkill
The file
We split this AGENTS.md into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
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
- Distinctive: line 13
Do not write tests for test helpers.
- Distinctive: line 15
If a helper needs its own tests, simplify it instead.
- Distinctive: line 20
Do not assert an upper bound on elapsed time.
- Distinctive: line 24
Test infrastructure defaults to a fake clock that never advances.
≈ “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.
Files
Agent — Agent Guide
Scoped guidance for the AI agent under Agent/ and its tests. Read this before
changing Agent code or Agent tests. When a rule here changes, update this guide in
the same PR.
Tests
- Build expected paths from the product's path constants (for example
HomePaths), not from string literals. A literal path can drift from what the product writes, and the assertion then passes for the wrong reason. - Prove that a new assertion fails when the condition it guards is broken.
- Do not write tests for test helpers. Keep fakes, harnesses and fake servers simple enough to verify by reading. Add a guard to a helper only when a real test needs it. If a helper needs its own tests, simplify it instead.
- Tests must not depend on machine speed. See "Timing in tests".
Timing in tests
- Do not assert an upper bound on elapsed time.
- Product code with a time budget takes a
TimeProvider(defaultTimeProvider.System). Its tests useFakeTimeProviderorSignalingFakeTimeProvider, and wait for the timer (WaitForTimerAsync) before callingAdvancewhen the timer may not be armed yet. Test infrastructure defaults to a fake clock that never advances. - A wait for something that must happen on a real thread, process or I/O uses the shared
60-second hang guard:
.WithAgentHangGuard(), orAgentTestTimeouts.HangGuardwhere aTimeSpanis needed (tests/Common/Utilities/AgentTestTimeouts.cs). - A short window of 50-100 ms is allowed only to show that something did not happen.
- A budget the test owns must not end the operation under test. Give the operation its own untimed token and assert which token cancelled it.
- A child process the test expects to be killed must outlive every wait in the test.
- Mark any test that is still speed-dependent
[Trait("Category", "LocalOnly")]. CI skips it.
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.