[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M79WN0VCKPHE1AD7T2K9M931":3,"skill-same:01M79WN0VCKPHE1AD7T2K9M931":57},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":30,"ownerDescription":31,"updatedAt":32,"install":33,"resources":46},"01M79WN0VCKPHE1AD7T2K9M931","code-testing",null,"Write or strengthen unit tests for existing code. Covers one function, class, or missing regression case up to a project-wide suite, in TUnit, xUnit, MSTest, NUnit, pytest, Vitest\u002FJest, Go, or another framework, including coverage-driven additions and strengthening tests an audit found weak. Focused work stays proportional; broad work runs a Research, Plan, Implement pipeline inline. DO NOT USE for repairing failing or flaky tests (systematic debugging owns that), only running tests, analysis-only audits, framework or platform migrations, or a behavior test blocked on a production seam that has not been selected.","Blind-Striker\u002Fagent-skills-and-plugins","Blind-Striker","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F1965259?v=4","01M79WFXTE5PH0C1VB7B54ANMN","Deniz İrgin","01M79WFXV5KZ5GST1K8G035DM5",1,"MIT","https:\u002F\u002Fgithub.com\u002FBlind-Striker\u002Fagent-skills-and-plugins\u002Fblob\u002Fmaster\u002Fcodex\u002Fdeniz-dotnet-general\u002Fskills\u002Fcode-testing\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: code-testing\ndescription: Write or strengthen unit tests for existing code. Covers one\n  function, class, or missing regression case up to a project-wide suite, in\n  TUnit, xUnit, MSTest, NUnit, pytest, Vitest\u002FJest, Go, or another framework,\n  including coverage-driven additions and strengthening tests an audit found\n  weak. Focused work stays proportional; broad work runs a Research, Plan,\n  Implement pipeline inline. DO NOT USE for repairing failing or flaky tests\n  (systematic debugging owns that), only running tests, analysis-only audits,\n  framework or platform migrations, or a behavior test blocked on a production\n  seam that has not been selected.\nlicense: MIT\n---\n\n# Code Testing Skill\n\nThe entry point for writing and strengthening tests for existing code. It\nhandles focused work directly and runs a Research → Plan → Implement pipeline in\nits own context for broad requests.\n\nThis skill does not repair failing or flaky tests. When the request is to fix a\nfailing, erroring, or flaky test, say that it is out of scope here, then load the\n`$deniz-process:systematic-debugging` skill and follow it instead. A failure in a\ntest this run generated is not such a request: fix that test as\n[Troubleshooting](#troubleshooting) describes.\n\n## Non-negotiable execution contract\n\nClassify scope **before editing**:\n\n- **Broad** (a project\u002Fpackage-wide suite, or multiple production\n  files\u002Fmodules): run the pipeline in [Step 3](#step-3-run-the-broad-pipeline-inline).\n  Create `research.md` and `plan.md` in a resolved non-stageable\n  `\u003CTESTAGENT_DIR>` before implementation, then `status.md` there after the\n  final test-quality review. If the state files are absent, the broad workflow\n  is incomplete.\n- **Focused** (the user explicitly limits work to one function\u002Fclass\u002Ffile or one\n  missing method): do not create intermediate state files. A sparse\n  project-wide request remains broad even when only one source module is\n  present.\n\nFor either scope, run the narrowest relevant test command to a clean exit.\nAlways apply [Report-safe test names and result validation](unit-test-generation.prompt.md#report-safe-test-names-and-result-validation),\nincluding when the caller supplies conventions. Preserve edge-case data and\nvalidate configured reports, not just console output.\nKeep the handoff proportional: for one to three focused requirements, use a\ncompact bullet list under a **Requirement coverage** label that names the tests\nand successful command; for broader or multi-requirement work, use a\n`Requirement | Evidence` table. Each requested behavior must cite an exact test\nname.\n\nBefore sending a broad-scope final response, check that the response itself\ncontains `| Requirement | Evidence |` and exact test names for every behavioral\nrow. A table in an internal plan is not enough. Do not summarize\naway those names into module-level bullets or an `Area | Tests` table.\n\nIntermediate state files are internal working data, never deliverables. Keep\n`\u003CTESTAGENT_DIR>` non-stageable, never place it or its files in\nversion-controlled workspace content, and never modify `.gitignore` to hide\nthem.\n\nTreat completeness as a requirement matrix, not a test-count target. Give every\nindependently requested state, boundary, error path, or interaction its own\nconcrete assertion. Combine cases only when one execution genuinely proves the\nwhole requested combination; do not let a parameterized happy-path case stand in\nfor an empty state, invalid discriminator, or before\u002Fat\u002Fafter boundary.\nFor broad requests that name several production modules or layers, give each\nnamed module direct tests for its non-trivial public behavior. Cross-module tests\nprove composition, but do not substitute for the requested module-level\ncoverage. Judge breadth by the behavior matrix, never by matching or exceeding a\nraw test count.\n\nFor broad work, research, plan, implementation, and review remain required.\nThey are phases of this one run, not separate agent calls.\n\nUse only capabilities available in the current runtime. Do not retry a missing\nskill under aliases or retry a policy-denied operation another way. Derive\nlanguage guidance from the repository's manifests and representative tests; do\nnot search installation directories for it.\nIf scratch storage is denied, keep the research and plan in context, continue\npermitted test edits, and report the missing state artifacts. If execution is\ndenied, continue permitted static review and report tests as unrun, never passed.\nNeither blocker authorizes modifying production code or weakening requirements.\n\nFor a **broad or comprehensive** request, the explicit matrix is the floor, not\nthe ceiling. Treat each requested module or layer as an inventory heading, not\none behavior: expand it into the bounded public operations and their distinct\nvalidation paths, branches, boundaries, interactions, and state transitions.\nAfter satisfying the explicit matrix, inspect each target API for observable\nequivalence partitions and invariants that the prompt did not name: identity,\nempty, singleton and representative interior inputs; exact boundaries plus an\nimmediately adjacent value; invalid partitions; and ordering, monotonicity,\nrollover, capacity, truncation, or state invariants implied by the implementation.\nAdd one mutation-relevant case per distinct partition not already proved, using\nparameterized or table-driven cases only for siblings that prove the same\nbehavior. A passing coverage threshold is validation, not a breadth stop\ncondition. Stop when remaining inputs exercise the same branch and invariant,\nnot merely when the explicit checklist is complete; never add cases only to\nraise the count.\n\n## When to Use This Skill\n\nUse this skill when you need to:\n\n- Generate unit tests for an entire project or specific files\n- Improve test coverage for existing codebases\n- Create test files that follow project conventions\n- Write tests that actually compile and pass\n- Add tests for new features or untested code\n- Generate or extend a C# suite. This estate is TUnit-first: when the test\n  project uses TUnit, or a new C# test project is needed, load the\n  `$deniz-dotnet-general:writing-tunit-tests` skill after this entry skill has\n  established scope and project conventions. An existing test project on\n  another framework keeps its framework.\n\n## When Not to Use\n\n- Repairing a failing, erroring, or flaky test (load the\n  `$deniz-process:systematic-debugging` skill)\n- Running or executing existing tests (use the `$deniz-dotnet-general:run-tests` skill)\n- Migrating between test frameworks\n- Answering a framework API or modernization question that does not ask to\n  generate tests\n- Analysis-only diagnosis of failing tests when no code or test changes are\n  requested\n\n## How It Works\n\nThis skill runs a **Research → Plan → Implement** pipeline in one context.\nResearch inventories the bounded targets and their conventions, Plan maps every\nrequirement to a concrete test, and Implement writes the tests, builds them,\nruns them, fixes them, and formats them. A focused request runs the same\nreasoning inline, without intermediate state files.\n\n## Step-by-Step Instructions\n\n### Step 1: Determine the user request\n\nClassify both intent and scope before editing:\n\n- **Generate or extend**: follow the generation workflow below.\n- **Audit then strengthen**: load the `$deniz-dotnet-general:test-anti-patterns` skill\n  and the `$deniz-dotnet-general:test-gap-analysis` skill for the existing tests in scope,\n  then add or strengthen tests only for the weaknesses and gaps they report.\n- **Repair**: out of scope. Load the `$deniz-process:systematic-debugging` skill\n  and follow it; do not edit tests or production code under this skill.\n\nMake sure you understand what user is asking and for what scope.\nWhen the user does not express strong requirements for test style, coverage goals, or conventions, source the guidelines from [unit-test-generation.prompt.md](unit-test-generation.prompt.md). This prompt provides best practices for discovering conventions, parameterization strategies, behavior-focused coverage, and language-specific patterns.\n\n### Step 2: Size the request before invoking anything\n\nMatch the machinery to the scope. Running the full pipeline on a one-file\nrequest costs turns and tool calls without improving the tests.\n\n| Scope | What it looks like | How to run it |\n| --- | --- | --- |\n| **Focused** | One function, class, or file; \"tests for X only\"; extending an existing suite with the missing cases | Skip intermediate state files. Keep the requirement checklist in your head (or in the final table), read only the target and one neighbouring test for conventions, write the tests, run the narrowest test command, review your own assertions inline. |\n| **Broad** | A project, package, or module set; \"comprehensive suite\"; a coverage threshold to clear across several files | Run the full Research → Plan → Implement pipeline in Step 3, with intermediate state files under `\u003CTESTAGENT_DIR>` and the completion contract below. |\n\nWhen in doubt, start focused and escalate only if the request turns out to span\nseveral files. Escalating costs one extra pass; running the broad pipeline on a\nfocused request costs several. A broad scope that fits one context runs each\nphase once. For a scope too large for one pass, extend the existing inventory\nand plan only for concrete remaining gaps; do not restart discovery.\n\nBefore ending a focused request, check all three conditions together:\n\n1. every named behavior has a concrete assertion, including each requested\n   boundary or error path;\n2. the narrow test command exited successfully;\n3. the final handoff maps those behaviors to exact test names and cites that\n   successful command.\n\nDo not replace requirement-level evidence with a generic list of covered areas.\n\n### Step 3: Run the broad pipeline inline\n\nExecute the Research → Plan → Implement sequence in this context, following the\n[unit-test-generation.prompt.md](unit-test-generation.prompt.md) guidelines.\nTreat the current workspace as authoritative even when it is sparse,\ngutted-looking, synthetic, or missing tracked files. Never restore or\nreconstruct it, including with `git checkout`, `git restore`, `git reset`, or\n`git clean`. If the workspace genuinely contains nothing testable, say so and\nstop.\n\nFor broad scope, resolve one absolute `\u003CTESTAGENT_DIR>` before creating\nintermediate state files:\n\n1. Prefer a host-provided session artifact or scratch directory.\n2. Otherwise, in a Git worktree run\n   `git rev-parse --path-format=absolute --git-path testagent`; this returns a\n   path in worktree-specific Git metadata that cannot be staged.\n3. Outside Git, create a unique directory under the operating system's\n   temporary directory.\n\nUse that absolute directory in every phase. The path may be inside the\nrepository's `.git` metadata directory, but it must not be version-controlled\nworkspace content, appear in `git status`, or be stageable.\n\n### Step 4: Execute with bounded context\n\nFor multi-file requests:\n\n1. Turn every explicit user requirement into a checklist before implementation. Include requested layers, collaborators to mock, boundary cases, integrations, coverage thresholds, and report artifacts. Copy multi-condition requirements verbatim — they must each map to one test that exercises the whole combination.\n2. Research only the requested module or project and write the checklist plus a compact target inventory to `\u003CTESTAGENT_DIR>\u002Fresearch.md`.\n3. Reuse manifests, symbol references, and deterministic pairing tools instead of reading every source and test file.\n4. For a substantial multi-file inventory, pair production sources with their existing tests once and record the suggested test paths in `\u003CTESTAGENT_DIR>\u002Fresearch.md`; do not repeat that discovery.\n5. Plan each target file once, then implement phases sequentially. Map every checklist item to at least one concrete test or explain why it is blocked.\n6. Build and test the narrow target during fix cycles. Run workspace-level\n   validation once at the end only for broad work, when the repository contract\n   requires that entry point, or when the changes can affect other projects.\n   Rebuild only after a concrete fix, at most three times for one diagnostic.\n   Stop when a diagnostic repeats without progress, a permission or toolchain\n   blocker is concrete, or the fix would leave the requested scope, and report\n   the work as `PARTIAL` with the exact command and diagnostic.\n7. Before reporting success, re-open the generated tests and verify every checklist item against concrete test names and assertions. Coverage alone is not evidence that a requested mock seam, boundary, state transition, or property combination was tested.\n8. When the repository has no representative tests, inspect the installed framework version, the project manifests, and the framework's primary documentation rather than inventing conventions.\n9. For .NET, classify SDK-style vs. classic non-SDK before choosing commands or creating files. In classic projects, preserve `packages.config`, existing framework\u002Fmock versions and custom base fixtures, add every new test file to the project's explicit `\u003CCompile Include>` items, and use the repository's MSBuild\u002Ftest-runner commands. Never modernize the project or dependency stack merely to generate tests. Load the `$deniz-dotnet-general:run-tests` skill for the repository-compatible test command.\n10. Confirm the run discovered the new tests. A zero-test run, or one whose\n    discovered count did not grow by the tests you added, validates nothing; in a\n    classic project, check the `\u003CCompile Include>` registration as well as the\n    test output.\n\n### Completion contract\n\nEvery scope must satisfy points 3–5 below. Points 1 and 2 are the **broad-scope**\nartifacts: on a focused request the same reasoning happens inline and no\nintermediate state files are written.\n\nDo not report completion until all of these are true:\n\n1. *(broad scope)* `\u003CTESTAGENT_DIR>\u002Fresearch.md` records the bounded target\n   inventory, existing test conventions, and the acceptance checklist.\n2. *(broad scope)* `\u003CTESTAGENT_DIR>\u002Fplan.md` maps each checklist item to a planned\n   test or an explicit blocker.\n3. Generated tests compile and pass with the narrowest relevant test command,\n   satisfying the shared report-safe naming and result-validation contract.\n   Report every skipped test and every setup or fixture failure; neither is a\n   pass.\n4. Every explicit user requirement is backed by a concrete test and assertion.\n   Fix missing mock seams, boundary cases, state transitions, and property\n   combinations even when coverage already passes. In the final summary, cite\n   at least one generated test name for every checklist item so completion is\n   auditable; if an item has no test to cite, keep implementing or report it as\n   blocked. For non-behavioral requirements such as scaffolding, scope limits,\n   commands, or coverage artifacts, cite the relevant file, command, or report\n   instead of forcing a test-name mapping.\n   A passing suite with fewer tests is not automatically weaker: judge\n   completeness by whether every independently requested behavior has direct,\n   nonredundant evidence, not by raw test volume.\n   For broad\u002Fcomprehensive scope, also verify that every observable equivalence\n   partition and invariant discovered in the bounded target APIs has one\n   mutation-relevant case, even when the prompt did not name it.\n   When the request names multiple modules, verify that each module's own\n   non-trivial public behavior has direct test evidence in addition to any\n   end-to-end composition test.\n5. Review the generated tests for behavior gaps and weak assertions. On a broad\n   scope, load the `$deniz-dotnet-general:test-gap-analysis` skill and the\n   `$deniz-dotnet-general:test-anti-patterns` skill, review the generated tests with them,\n   and record the findings and fixes in `\u003CTESTAGENT_DIR>\u002Fstatus.md`. On a focused scope,\n   do the equivalent review inline — re-read each generated assertion against\n   the source — without spawning extra passes.\n\nLead the final response with `SUCCESS` only when every required validation\npassed; otherwise lead with `PARTIAL` or `BLOCKED` and give the exact command\nand diagnostic for each unrun or failed validation.\n\nThe final response must provide requirement-by-requirement evidence. Use compact\nbullets under a **Requirement coverage** label for one to three focused\nrequirements; use a `Requirement | Evidence` table for broader scopes.\nBehavioral evidence cites exact generated test names. Non-behavioral evidence\ncites the relevant project file, validation command, or coverage report. A\ngeneric list of tested areas is not a substitute.\n\nPreserve the user's exact meaning in each evidence item; quote verbatim only\nwhen wording distinguishes a required combination. A test that merely exercises\nthe same collaborators does not satisfy a requirement about their interaction,\nand per-class requirements need a citation per class.\n\n**Cite a clean run, not an attempt.** The commands behind the final evidence must\nhave finished successfully: quote the final passing test summary and, when\nthresholds were requested, the per-module coverage table from a run that exited\n0. If the last coverage run exited non-zero, fix it and re-run before reporting;\nnever infer threshold clearance from a failed or partial run.\n\nBefore reporting, inspect the final working-tree changes and confirm that\n`research.md`, `plan.md`, `status.md`, and any other intermediate state files are\nnot among the changes intended for commit.\n\n## State Management\n\nBroad-scope runs store intermediate state files in a non-stageable\n`\u003CTESTAGENT_DIR>` backed by host scratch storage, Git metadata, or OS temp. A\nfocused request does not create these files:\n\n| File                     | Purpose                      |\n| ------------------------ | ---------------------------- |\n| `\u003CTESTAGENT_DIR>\u002Fresearch.md` | Codebase analysis results    |\n| `\u003CTESTAGENT_DIR>\u002Fplan.md`     | Phased implementation plan   |\n| `\u003CTESTAGENT_DIR>\u002Fstatus.md`   | Final quality review and fixes |\n\n## Requirements\n\n- Project must have a build\u002Ftest system configured\n- Testing framework should be installed (or installable)\n\nClassic non-SDK .NET projects are supported when their existing build\u002Ftest\ntoolchain is available. When it is not available on the current machine, the\nagent can still add and register version-compatible tests, but must report\nexecution as blocked rather than substituting `dotnet test`.\n\n## Troubleshooting\n\n### Tests don't compile\n\nRead the exact compiler error, the project manifest, and one working neighboring\ntest before changing generated code. Check `\u003CTESTAGENT_DIR>\u002Fplan.md` for the\nexpected test structure, make the smallest version-compatible correction, then\nrebuild the narrow target. Do not modernize packages or the project shape merely\nto make the generated tests compile.\n\n### Generated tests fail\n\nMost failures in generated tests are caused by **wrong expected values in assertions**, not production code bugs:\n\n1. Read the actual test output\n2. Read the production code to understand correct behavior\n3. Fix the assertion, not the production code\n4. Never mark tests `[Ignore]` or `[Skip]` just to make them pass\n\n### Wrong testing framework detected\n\nSpecify your preferred framework in the initial request: \"Generate Jest tests for...\"\n\n### Environment-dependent tests fail\n\nTests that depend on external services, network endpoints, specific ports, or precise timing will fail in CI environments. Focus on unit tests with mocked dependencies instead.\n\n### Broader validation fails\n\nDuring implementation, build and test the narrow target. Run a solution or\nworkspace-level command only for broad work, when the repository contract uses\nthat entry point, or when the targeted change can affect other projects. Do not\nturn a focused test request into an unconditional full non-incremental build.\n","# Code Testing Skill\n\nThe entry point for writing and strengthening tests for existing code. It\nhandles focused work directly and runs a Research → Plan → Implement pipeline in\nits own context for broad requests.\n\nThis skill does not repair failing or flaky tests. When the request is to fix a\nfailing, erroring, or flaky test, say that it is out of scope here, then load the\n`$deniz-process:systematic-debugging` skill and follow it instead. A failure in a\ntest this run generated is not such a request: fix that test as\n[Troubleshooting](#troubleshooting) describes.\n\n## Non-negotiable execution contract\n\nClassify scope **before editing**:\n\n- **Broad** (a project\u002Fpackage-wide suite, or multiple production\n  files\u002Fmodules): run the pipeline in [Step 3](#step-3-run-the-broad-pipeline-inline).\n  Create `research.md` and `plan.md` in a resolved non-stageable\n  `\u003CTESTAGENT_DIR>` before implementation, then `status.md` there after the\n  final test-quality review. If the state files are absent, the broad workflow\n  is incomplete.\n- **Focused** (the user explicitly limits work to one function\u002Fclass\u002Ffile or one\n  missing method): do not create intermediate state files. A sparse\n  project-wide request remains broad even when only one source module is\n  present.\n\nFor either scope, run the narrowest relevant test command to a clean exit.\nAlways apply [Report-safe test names and result validation](unit-test-generation.prompt.md#report-safe-test-names-and-result-validation),\nincluding when the caller supplies conventions. Preserve edge-case data and\nvalidate configured reports, not just console output.\nKeep the handoff proportional: for one to three focused requirements, use a\ncompact bullet list under a **Requirement coverage** label that names the tests\nand successful command; for broader or multi-requirement work, use a\n`Requirement | Evidence` table. Each requested behavior must cite an exact test\nname.\n\nBefore sending a broad-scope final response, check that the response itself\ncontains `| Requirement | Evidence |` and exact test names for every behavioral\nrow. A table in an internal plan is not enough. Do not summarize\naway those names into module-level bullets or an `Area | Tests` table.\n\nIntermediate state files are internal working data, never deliverables. Keep\n`\u003CTESTAGENT_DIR>` non-stageable, never place it or its files in\nversion-controlled workspace content, and never modify `.gitignore` to hide\nthem.\n\nTreat completeness as a requirement matrix, not a test-count target. Give every\nindependently requested state, boundary, error path, or interaction its own\nconcrete assertion. Combine cases only when one execution genuinely proves the\nwhole requested combination; do not let a parameterized happy-path case stand in\nfor an empty state, invalid discriminator, or before\u002Fat\u002Fafter boundary.\nFor broad requests that name several production modules or layers, give each\nnamed module direct tests for its non-trivial public behavior. Cross-module tests\nprove composition, but do not substitute for the requested module-level\ncoverage. Judge breadth by the behavior matrix, never by matching or exceeding a\nraw test count.\n\nFor broad work, research, plan, implementation, and review remain required.\nThey are phases of this one run, not separate agent calls.\n\nUse only capabilities available in the current runtime. Do not retry a missing\nskill under aliases or retry a policy-denied operation another way. Derive\nlanguage guidance from the repository's manifests and representative tests; do\nnot search installation directories for it.\nIf scratch storage is denied, keep the research and plan in context, continue\npermitted test edits, and report the missing state artifacts. If execution is\ndenied, continue permitted static review and report tests as unrun, never passed.\nNeither blocker authorizes modifying production code or weakening requirements.\n\nFor a **broad or comprehensive** request, the explicit matrix is the floor, not\nthe ceiling. Treat each requested module or layer as an inventory heading, not\none behavior: expand it into the bounded public operations and their distinct\nvalidation paths, branches, boundaries, interactions, and state transitions.\nAfter satisfying the explicit matrix, inspect each target API for observable\nequivalence partitions and invariants that the prompt did not name: identity,\nempty, singleton and representative interior inputs; exact boundaries plus an\nimmediately adjacent value; invalid partitions; and ordering, monotonicity,\nrollover, capacity, truncation, or state invariants implied by the implementation.\nAdd one mutation-relevant case per distinct partition not already proved, using\nparameterized or table-driven cases only for siblings that prove the same\nbehavior. A passing coverage threshold is validation, not a breadth stop\ncondition. Stop when remaining inputs exercise the same branch and invariant,\nnot merely when the explicit checklist is complete; never add cases only to\nraise the count.\n\n## When to Use This Skill\n\nUse this skill when you need to:\n\n- Generate unit tests for an entire project or specific files\n- Improve test coverage for existing codebases\n- Create test files that follow project conventions\n- Write tests that actually compile and pass\n- Add tests for new features or untested code\n- Generate or extend a C# suite. This estate is TUnit-first: when the test\n  project uses TUnit, or a new C# test project is needed, load the\n  `$deniz-dotnet-general:writing-tunit-tests` skill after this entry skill has\n  established scope and project conventions. An existing test project on\n  another framework keeps its framework.\n\n## When Not to Use\n\n- Repairing a failing, erroring, or flaky test (load the\n  `$deniz-process:systematic-debugging` skill)\n- Running or executing existing tests (use the `$deniz-dotnet-general:run-tests` skill)\n- Migrating between test frameworks\n- Answering a framework API or modernization question that does not ask to\n  generate tests\n- Analysis-only diagnosis of failing tests when no code or test changes are\n  requested\n\n## How It Works\n\nThis skill runs a **Research → Plan → Implement** pipeline in one context.\nResearch inventories the bounded targets and their conventions, Plan maps every\nrequirement to a concrete test, and Implement writes the tests, builds them,\nruns them, fixes them, and formats them. A focused request runs the same\nreasoning inline, without intermediate state files.\n\n## Step-by-Step Instructions\n\n### Step 1: Determine the user request\n\nClassify both intent and scope before editing:\n\n- **Generate or extend**: follow the generation workflow below.\n- **Audit then strengthen**: load the `$deniz-dotnet-general:test-anti-patterns` skill\n  and the `$deniz-dotnet-general:test-gap-analysis` skill for the existing tests in scope,\n  then add or strengthen tests only for the weaknesses and gaps they report.\n- **Repair**: out of scope. Load the `$deniz-process:systematic-debugging` skill\n  and follow it; do not edit tests or production code under this skill.\n\nMake sure you understand what user is asking and for what scope.\nWhen the user does not express strong requirements for test style, coverage goals, or conventions, source the guidelines from [unit-test-generation.prompt.md](unit-test-generation.prompt.md). This prompt provides best practices for discovering conventions, parameterization strategies, behavior-focused coverage, and language-specific patterns.\n\n### Step 2: Size the request before invoking anything\n\nMatch the machinery to the scope. Running the full pipeline on a one-file\nrequest costs turns and tool calls without improving the tests.\n\n| Scope | What it looks like | How to run it |\n| --- | --- | --- |\n| **Focused** | One function, class, or file; \"tests for X only\"; extending an existing suite with the missing cases | Skip intermediate state files. Keep the requirement checklist in your head (or in the final table), read only the target and one neighbouring test for conventions, write the tests, run the narrowest test command, review your own assertions inline. |\n| **Broad** | A project, package, or module set; \"comprehensive suite\"; a coverage threshold to clear across several files | Run the full Research → Plan → Implement pipeline in Step 3, with intermediate state files under `\u003CTESTAGENT_DIR>` and the completion contract below. |\n\nWhen in doubt, start focused and escalate only if the request turns out to span\nseveral files. Escalating costs one extra pass; running the broad pipeline on a\nfocused request costs several. A broad scope that fits one context runs each\nphase once. For a scope too large for one pass, extend the existing inventory\nand plan only for concrete remaining gaps; do not restart discovery.\n\nBefore ending a focused request, check all three conditions together:\n\n1. every named behavior has a concrete assertion, including each requested\n   boundary or error path;\n2. the narrow test command exited successfully;\n3. the final handoff maps those behaviors to exact test names and cites that\n   successful command.\n\nDo not replace requirement-level evidence with a generic list of covered areas.\n\n### Step 3: Run the broad pipeline inline\n\nExecute the Research → Plan → Implement sequence in this context, following the\n[unit-test-generation.prompt.md](unit-test-generation.prompt.md) guidelines.\nTreat the current workspace as authoritative even when it is sparse,\ngutted-looking, synthetic, or missing tracked files. Never restore or\nreconstruct it, including with `git checkout`, `git restore`, `git reset`, or\n`git clean`. If the workspace genuinely contains nothing testable, say so and\nstop.\n\nFor broad scope, resolve one absolute `\u003CTESTAGENT_DIR>` before creating\nintermediate state files:\n\n1. Prefer a host-provided session artifact or scratch directory.\n2. Otherwise, in a Git worktree run\n   `git rev-parse --path-format=absolute --git-path testagent`; this returns a\n   path in worktree-specific Git metadata that cannot be staged.\n3. Outside Git, create a unique directory under the operating system's\n   temporary directory.\n\nUse that absolute directory in every phase. The path may be inside the\nrepository's `.git` metadata directory, but it must not be version-controlled\nworkspace content, appear in `git status`, or be stageable.\n\n### Step 4: Execute with bounded context\n\nFor multi-file requests:\n\n1. Turn every explicit user requirement into a checklist before implementation. Include requested layers, collaborators to mock, boundary cases, integrations, coverage thresholds, and report artifacts. Copy multi-condition requirements verbatim — they must each map to one test that exercises the whole combination.\n2. Research only the requested module or project and write the checklist plus a compact target inventory to `\u003CTESTAGENT_DIR>\u002Fresearch.md`.\n3. Reuse manifests, symbol references, and deterministic pairing tools instead of reading every source and test file.\n4. For a substantial multi-file inventory, pair production sources with their existing tests once and record the suggested test paths in `\u003CTESTAGENT_DIR>\u002Fresearch.md`; do not repeat that discovery.\n5. Plan each target file once, then implement phases sequentially. Map every checklist item to at least one concrete test or explain why it is blocked.\n6. Build and test the narrow target during fix cycles. Run workspace-level\n   validation once at the end only for broad work, when the repository contract\n   requires that entry point, or when the changes can affect other projects.\n   Rebuild only after a concrete fix, at most three times for one diagnostic.\n   Stop when a diagnostic repeats without progress, a permission or toolchain\n   blocker is concrete, or the fix would leave the requested scope, and report\n   the work as `PARTIAL` with the exact command and diagnostic.\n7. Before reporting success, re-open the generated tests and verify every checklist item against concrete test names and assertions. Coverage alone is not evidence that a requested mock seam, boundary, state transition, or property combination was tested.\n8. When the repository has no representative tests, inspect the installed framework version, the project manifests, and the framework's primary documentation rather than inventing conventions.\n9. For .NET, classify SDK-style vs. classic non-SDK before choosing commands or creating files. In classic projects, preserve `packages.config`, existing framework\u002Fmock versions and custom base fixtures, add every new test file to the project's explicit `\u003CCompile Include>` items, and use the repository's MSBuild\u002Ftest-runner commands. Never modernize the project or dependency stack merely to generate tests. Load the `$deniz-dotnet-general:run-tests` skill for the repository-compatible test command.\n10. Confirm the run discovered the new tests. A zero-test run, or one whose\n    discovered count did not grow by the tests you added, validates nothing; in a\n    classic project, check the `\u003CCompile Include>` registration as well as the\n    test output.\n\n### Completion contract\n\nEvery scope must satisfy points 3–5 below. Points 1 and 2 are the **broad-scope**\nartifacts: on a focused request the same reasoning happens inline and no\nintermediate state files are written.\n\nDo not report completion until all of these are true:\n\n1. *(broad scope)* `\u003CTESTAGENT_DIR>\u002Fresearch.md` records the bounded target\n   inventory, existing test conventions, and the acceptance checklist.\n2. *(broad scope)* `\u003CTESTAGENT_DIR>\u002Fplan.md` maps each checklist item to a planned\n   test or an explicit blocker.\n3. Generated tests compile and pass with the narrowest relevant test command,\n   satisfying the shared report-safe naming and result-validation contract.\n   Report every skipped test and every setup or fixture failure; neither is a\n   pass.\n4. Every explicit user requirement is backed by a concrete test and assertion.\n   Fix missing mock seams, boundary cases, state transitions, and property\n   combinations even when coverage already passes. In the final summary, cite\n   at least one generated test name for every checklist item so completion is\n   auditable; if an item has no test to cite, keep implementing or report it as\n   blocked. For non-behavioral requirements such as scaffolding, scope limits,\n   commands, or coverage artifacts, cite the relevant file, command, or report\n   instead of forcing a test-name mapping.\n   A passing suite with fewer tests is not automatically weaker: judge\n   completeness by whether every independently requested behavior has direct,\n   nonredundant evidence, not by raw test volume.\n   For broad\u002Fcomprehensive scope, also verify that every observable equivalence\n   partition and invariant discovered in the bounded target APIs has one\n   mutation-relevant case, even when the prompt did not name it.\n   When the request names multiple modules, verify that each module's own\n   non-trivial public behavior has direct test evidence in addition to any\n   end-to-end composition test.\n5. Review the generated tests for behavior gaps and weak assertions. On a broad\n   scope, load the `$deniz-dotnet-general:test-gap-analysis` skill and the\n   `$deniz-dotnet-general:test-anti-patterns` skill, review the generated tests with them,\n   and record the findings and fixes in `\u003CTESTAGENT_DIR>\u002Fstatus.md`. On a focused scope,\n   do the equivalent review inline — re-read each generated assertion against\n   the source — without spawning extra passes.\n\nLead the final response with `SUCCESS` only when every required validation\npassed; otherwise lead with `PARTIAL` or `BLOCKED` and give the exact command\nand diagnostic for each unrun or failed validation.\n\nThe final response must provide requirement-by-requirement evidence. Use compact\nbullets under a **Requirement coverage** label for one to three focused\nrequirements; use a `Requirement | Evidence` table for broader scopes.\nBehavioral evidence cites exact generated test names. Non-behavioral evidence\ncites the relevant project file, validation command, or coverage report. A\ngeneric list of tested areas is not a substitute.\n\nPreserve the user's exact meaning in each evidence item; quote verbatim only\nwhen wording distinguishes a required combination. A test that merely exercises\nthe same collaborators does not satisfy a requirement about their interaction,\nand per-class requirements need a citation per class.\n\n**Cite a clean run, not an attempt.** The commands behind the final evidence must\nhave finished successfully: quote the final passing test summary and, when\nthresholds were requested, the per-module coverage table from a run that exited\n0. If the last coverage run exited non-zero, fix it and re-run before reporting;\nnever infer threshold clearance from a failed or partial run.\n\nBefore reporting, inspect the final working-tree changes and confirm that\n`research.md`, `plan.md`, `status.md`, and any other intermediate state files are\nnot among the changes intended for commit.\n\n## State Management\n\nBroad-scope runs store intermediate state files in a non-stageable\n`\u003CTESTAGENT_DIR>` backed by host scratch storage, Git metadata, or OS temp. A\nfocused request does not create these files:\n\n| File                     | Purpose                      |\n| ------------------------ | ---------------------------- |\n| `\u003CTESTAGENT_DIR>\u002Fresearch.md` | Codebase analysis results    |\n| `\u003CTESTAGENT_DIR>\u002Fplan.md`     | Phased implementation plan   |\n| `\u003CTESTAGENT_DIR>\u002Fstatus.md`   | Final quality review and fixes |\n\n## Requirements\n\n- Project must have a build\u002Ftest system configured\n- Testing framework should be installed (or installable)\n\nClassic non-SDK .NET projects are supported when their existing build\u002Ftest\ntoolchain is available. When it is not available on the current machine, the\nagent can still add and register version-compatible tests, but must report\nexecution as blocked rather than substituting `dotnet test`.\n\n## Troubleshooting\n\n### Tests don't compile\n\nRead the exact compiler error, the project manifest, and one working neighboring\ntest before changing generated code. Check `\u003CTESTAGENT_DIR>\u002Fplan.md` for the\nexpected test structure, make the smallest version-compatible correction, then\nrebuild the narrow target. Do not modernize packages or the project shape merely\nto make the generated tests compile.\n\n### Generated tests fail\n\nMost failures in generated tests are caused by **wrong expected values in assertions**, not production code bugs:\n\n1. Read the actual test output\n2. Read the production code to understand correct behavior\n3. Fix the assertion, not the production code\n4. Never mark tests `[Ignore]` or `[Skip]` just to make them pass\n\n### Wrong testing framework detected\n\nSpecify your preferred framework in the initial request: \"Generate Jest tests for...\"\n\n### Environment-dependent tests fail\n\nTests that depend on external services, network endpoints, specific ports, or precise timing will fail in CI environments. Focus on unit tests with mocked dependencies instead.\n\n### Broader validation fails\n\nDuring implementation, build and test the narrow target. Run a solution or\nworkspace-level command only for broad work, when the repository contract uses\nthat entry point, or when the targeted change can affect other projects. Do not\nturn a focused test request into an unconditional full non-incremental build.\n",{"name":5,"license":15,"description":7},"master","b30b1c1908dbc56a8184128f18803053d762ccb5","codex\u002Fdeniz-dotnet-general\u002Fskills\u002Fcode-testing",0,"https:\u002F\u002Fgithub.com\u002FBlind-Striker\u002Fagent-skills-and-plugins","https:\u002F\u002Fgithub.com\u002FBlind-Striker","denizirgin.com","VP of Engineering - \r\n@armutcom @homeruntech @prontopro\r\n |  Speaker \r\n@CodeFiction\r\n, http:\u002F\u002Fcodefiction.tech | Developer, gamer, blogger, maiden iron fan","2026-11-13 00:23:18",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add Blind-Striker\u002Fagent-skills-and-plugins --skill code-testing",{"tool":39,"command":40},"bunx","bunx skills add Blind-Striker\u002Fagent-skills-and-plugins --skill code-testing",{"tool":42,"command":43},"pnpm","pnpm dlx skills add Blind-Striker\u002Fagent-skills-and-plugins --skill code-testing","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FBlind-Striker\u002Fagent-skills-and-plugins\u002Fblob\u002Fmaster\u002Fcodex\u002Fdeniz-dotnet-general\u002Fskills\u002Fcode-testing\u002FSKILL.md",true,[47,53],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"agents\u002Fopenai.yaml","REFERENCE","参考文档",232,"https:\u002F\u002Fgithub.com\u002FBlind-Striker\u002Fagent-skills-and-plugins\u002Fblob\u002Fmaster\u002Fcodex\u002Fdeniz-dotnet-general\u002Fskills\u002Fcode-testing\u002Fagents\u002Fopenai.yaml",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"unit-test-generation.prompt.md",14211,"https:\u002F\u002Fgithub.com\u002FBlind-Striker\u002Fagent-skills-and-plugins\u002Fblob\u002Fmaster\u002Fcodex\u002Fdeniz-dotnet-general\u002Fskills\u002Fcode-testing\u002Funit-test-generation.prompt.md",[58,65,72,79,86,93],{"skillNo":59,"name":60,"nameZh":6,"description":61,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"publishedAt":19},"01M79YGWBJ52PHMSTKJ3RJ3AWR","writing-tunit-tests","Write tests in TUnit — the attributes, the async assertion model, data-driven cases, lifecycle hooks, and the parallelism rules. USE FOR writing or fixing a test in a TUnit project, translating a test idea into TUnit syntax, or reviewing TUnit test code for API mistakes. DO NOT USE FOR deciding what to test, judging whether a test is any good, or running tests.","https:\u002F\u002Fgithub.com\u002FBlind-Striker\u002Fagent-skills-and-plugins\u002Fblob\u002Fmaster\u002Fskills\u002Fdeniz-dotnet-general\u002Fwriting-tunit-tests\u002FSKILL.md",[],[],{"skillNo":66,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"publishedAt":19},"01M79YGR3Y9RCRS6TD71D0KKP8","dotnet-walls","Walls for a .NET change in a repository without its own quality gates — pinned, ephemeral analyzer, format, clone and scanner commands for the verify step, their identity ratchet, and the build traps. Use during a walls preflight, or when a worker's verify step needs .NET quality gates.","https:\u002F\u002Fgithub.com\u002FBlind-Striker\u002Fagent-skills-and-plugins\u002Fblob\u002Fmaster\u002Fskills\u002Fdeniz-dotnet-general\u002Fdotnet-walls\u002FSKILL.md",[],[],{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":19},"01M79YGFNJBST7D803NQW3SBYM","csharp-15","Write, review and modernize C# 15 (.NET 11) code correctly. Covers every C# 15 feature - collection expression arguments (`[with(...), ...]`), union types (`union`, `[Union]`, `IUnion`, custom unions), closed class hierarchies (`closed`), extension indexers, labeled `break`\u002F`continue`, non-virtual static interface members on runtimes without default interface methods, and the preview memory-safety model (`unsafe(expr)`, `safe`, requires-unsafe members). Use when code targets net11.0 or LangVersion 15\u002Fpreview, when the user mentions any of these features, when replacing marker interfaces, OneOf-style results, `goto`\u002Fflag loop exits or capacity\u002Fcomparer boilerplate, or when the build reports CS8121, CS8509, CS8655, CS8510, CS0121\u002FCS0411 on a `with(...)` element, CS9370, CS9380 or CS9363.","https:\u002F\u002Fgithub.com\u002FBlind-Striker\u002Fagent-skills-and-plugins\u002Fblob\u002Fmaster\u002Fskills\u002Fdeniz-dotnet-general\u002Fcsharp-15\u002FSKILL.md",[],[],{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":19},"01M79YGAV031473HSX6FN87ENH","verified-waves","Run one task as verified multi-agent waves. Plans the fan-out, sends findings to fresh skeptics, keeps a ledger, and reports N of M done; small tasks stay solo.","https:\u002F\u002Fgithub.com\u002FBlind-Striker\u002Fagent-skills-and-plugins\u002Fblob\u002Fmaster\u002Fskills\u002Fdeniz-coordination\u002Fverified-waves\u002FSKILL.md",[],[],{"skillNo":87,"name":88,"nameZh":6,"description":89,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":90,"categories":91,"keywords":92,"publishedAt":19},"01M79YG6QXNND4WRPP3GATSY8J","harness","The bundled dispatch harness, which runs a task file on worker models through OpenCode or Codex, fences their writes, escalates their asks to the operator, settles each result from git and its verify commands, and records every step in a ledger. Its commands, task file, modes and write fence, configuration, runs and their recovery, ledger and exit codes.","https:\u002F\u002Fgithub.com\u002FBlind-Striker\u002Fagent-skills-and-plugins\u002Fblob\u002Fmaster\u002Fskills\u002Fdeniz-coordination\u002Fharness\u002FSKILL.md",[],[],{"skillNo":94,"name":95,"nameZh":6,"description":96,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"publishedAt":19},"01M79YG2P5S91SDTA5912XJ4FV","delegating-to-worker-models","Use when you hand off implementation, a plan, research a decision will rely on, or a review, and you brief the worker and check what comes back. A subagent that searches, reads in parallel, or keeps your context small is ordinary tool use. Covers what stays with you, the mechanism, assigning work by strength, the walls preflight, prompts without contradictions, the worker contract and model profiles, evidence from your own run, review, and the fix-up loop.","https:\u002F\u002Fgithub.com\u002FBlind-Striker\u002Fagent-skills-and-plugins\u002Fblob\u002Fmaster\u002Fskills\u002Fdeniz-coordination\u002Fdelegating-to-worker-models\u002FSKILL.md",[],[]]