testing-verification
Skill0 starsChanged 3 days agoMITView on GitHub
Software testing and independent verification standard. Requires that tests are genuinely executed, results are inspectable, failures are not hidden, and the verification scope is explicit.
3 other projects use a version of this.
Adopted by
Most recently taken up by
- AL-JANEF/janef-forgetheir own version, first seen Sep 19, 2026
- AL-JANEF/janefskillstheir own version, first seen Sep 19, 2026
- thienanblog/awesome-ai-agent-skillstheir own version, first seen May 3, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- thienanblog/awesome-ai-agent-skills★ 68same name
- AL-JANEF/janef-forge★ 1same name
- AL-JANEF/janefskills★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository17 other files it gives its agents.
- .cursor/rules/engineering-context.mdcCursor rule
- change-report-deliverySkill
- code-asset-inventorySkill
- data-audit-reproducibilitySkill
- data-first-executionSkill
- documentation-structureSkill
- engineering-context-completenessSkill
- engineering-transparencySkill
- explicit-instruction-fidelitySkill
- interface-contractSkill
- language-expressionSkill
- production-loggingSkill
- repository-publication-hygieneSkill
- task-execution-governanceSkill
- technical-documentationSkill
- unfinished-work-trackingSkill
- user-intent-prioritySkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
15 rules. 0 are common, 15 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
15
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
- Distinctive: line 25
Verify the input, processing, and output of a single function or component.
- Distinctive: line 29
Verify the data transfer and call relationships between multiple modules.
- Distinctive: line 33
Verify request format, response structure, field types, and exceptional behavior.
- Distinctive: line 37
Verify data counts, fields, types, null values, uniqueness, and business constraints.
- Distinctive: line 41
Verify that the current change did not break existing functionality.
- Distinctive: line 45
Verify that a complete business flow runs correctly.
- Distinctive: line 51
Every significant test must state:
- Distinctive: line 61
Test data must be clearly distinguished as:
- Distinctive: line 102
When a test fails:
- Distinctive: line 105
Check the failure location.
- Distinctive: line 106
Check the input data.
- Distinctive: line 107
Check the actual output.
- Distinctive: line 114
If the test itself is problematic, this must be clearly stated and the test basis corrected.
- Distinctive: line 131
Prefer an independent test directory, a temporary database, or an isolated environment.
- Distinctive: line 133
After testing, clearly state:
“Common” is measured: kept by at least 6 in 10 of the 3 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.
Files
Testing and Verification Standard
I. Core Principle
The completion of code writing does not mean the functionality is correct.
The ability of a program to run does not mean the result is correct.
The existence of test code does not mean the test has been executed.
A passing test must be based on real execution results.
II. Test Types
Select according to the task:
1. Unit Tests
Verify the input, processing, and output of a single function or component.
2. Integration Tests
Verify the data transfer and call relationships between multiple modules.
3. Interface Tests
Verify request format, response structure, field types, and exceptional behavior.
4. Data Quality Tests
Verify data counts, fields, types, null values, uniqueness, and business constraints.
5. Regression Tests
Verify that the current change did not break existing functionality.
6. End-to-End Tests
Verify that a complete business flow runs correctly.
Executing every test type irrelevant to the current task purely for form is not acceptable.
III. Must Be Clarified Before Testing
Every significant test must state:
- Test objective.
- Test target.
- Test input.
- Test data source.
- Expected result.
- Pass criteria.
- Failure criteria.
Test data must be clearly distinguished as:
- Real production data.
- Masked data.
- Manually constructed test data.
- Mock data.
Mock results must not be described as real production environment results.
IV. Test Execution Requirements
Tests must actually be executed.
Record:
- Test command.
- Execution time.
- Environment used.
- Data used.
- Input parameters.
- Actual output.
- Exit status.
- Failure messages.
- Test report location.
Claiming a test passed based only on reading the test code is not acceptable.
V. Test Result Classification
Must clearly distinguish:
- PASS: actually executed and passed.
- FAIL: actually executed but failed.
- BLOCKED: could not be executed due to environment or dependency issues.
- NOT_RUN: not executed.
- NOT_APPLICABLE: not applicable to the current task.
Writing BLOCKED or NOT_RUN as PASS is forbidden.
VI. Failure Handling
When a test fails:
- Preserve the original error.
- Check the failure location.
- Check the input data.
- Check the actual output.
- Analyze the cause of the failure.
- Fix the related problem.
- Re-run the necessary tests.
Passing results must not be manufactured by deleting failing tests, skipping tests, or changing expected values.
If the test itself is problematic, this must be clearly stated and the test basis corrected.
VII. Test Coverage Scope
Determine according to the impact scope of the change:
- Single function change: test the related function.
- Module change: test the module and its direct dependencies.
- Interface change: test caller compatibility.
- Cross-module change: test the complete data chain.
- Data processing change: test data counts and field changes.
- Logging change: inspect the actually generated logs.
VIII. Test Data Protection
Tests must not modify or delete real production data without authorization.
Prefer an independent test directory, a temporary database, or an isolated environment.
After testing, clearly state:
- Whether files were produced.
- Whether data was modified.
- Whether temporary resources were cleaned up.
- Whether test results were retained.
IX. Independent Review
For significant changes, a review separated from the implementation process should be conducted.
Review focus:
- Whether the original requirement is satisfied.
- Whether new problems were introduced.
- Whether boundary conditions were missed.
- Whether existing interfaces were affected.
- Whether untested paths exist.
- Whether data inconsistency exists.
- Whether unnecessary duplicate implementation exists.
Review conclusions must have a basis.
X. Final Delivery
State:
- Which tests were actually executed.
- Which tests passed.
- Which tests failed.
- Which tests were not executed.
- Why they were not executed.
- What risks remain.
- Where the test results are stored.
XI. Final Principle
Testing verifies actual behavior; it is not a means of finding support for the Agent's claims of completion.
Without verification, do not claim that verification has been performed.
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.

