테스트·보안 스킬

Guides authorized Active Directory security testing with scope boundaries, alternative testing directions, and evidence requirements. Helps agents document domain permissions, reproduction records, and observed impact before deciding whether findings are supported or testing should move elsewhere.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Defining the scope of Active Directory tests
• Reviewing stalled privilege escalation tests
• Organizing domain test evidence

Guide feature changes and bug fixes through test-driven development: write a failing test, add the minimum code to pass, then improve the structure. Check coverage and include tests for boundary conditions, errors, and relevant integration points.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Adding features with tests written first
• Writing a regression test before fixing a bug
• Keeping tests passing while improving code structure

Check whether work is ready to call complete by running relevant tests, builds, or other checks. Review fresh output and compare results with requirements, then report completion or remaining problems with supporting evidence.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Checking test and build results before reporting completion
• Confirming that a fix resolves the original bug
• Verifying changes reported as complete by another agent

Check code and agent outputs, track reliability scores, and roll back changes that fail a chosen threshold. Includes quality reports, live monitoring, and integration with automated build and deployment workflows.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Checking code produced by AI agents
• Rolling back changes that fail verification
• Tracking quality trends over time

Analyze a codebase to identify valuable assets, attack entry points, and realistic abuse paths. Produce a concise Markdown threat model that explains risk priorities, distinguishes existing protections from proposed mitigations, and makes assumptions explicit.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Map attack entry points in a codebase
• Assess threats within a specific project directory
• Prioritize risks to sensitive data and credentials

Analyze git history to map contributors to security-sensitive files, identify reliance on too few people, and find stale areas. Export CSV and JSON results for ownership reviews, graph exploration, and comparison with CODEOWNERS.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Find stale security-sensitive code with few maintainers
• Check reliance on individual authentication or cryptography maintainers
• Compare CODEOWNERS with actual contribution history

Review code for security risks in authentication, APIs, input handling, and data access. Identify issues such as exposed secrets, injection, and missing access checks, then report severity, file evidence, proposed fixes, and verification steps.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Reviewing authentication and access controls
• Checking for exposed secrets and unsafe logs
• Auditing input handling and file uploads

Review Python, JavaScript/TypeScript, and Go code for security issues using guidance for the project's languages and frameworks. Get prioritized findings, suggested improvements, or help writing new code with safer defaults.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Review an existing application's security
• Create a report organized by severity
• Write new code with safer defaults

Build tests with fast-check for TypeScript and JavaScript or Hypothesis for Python. Generate varied inputs to check rules code should always satisfy, discover edge cases, and simplify failing inputs into small examples for debugging.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Checking that encoding and decoding restore the original data
• Testing rules that sorting must preserve
• Generating inputs with complex constraints

Measure indicator calculation speed with BenchmarkDotNet, a benchmarking tool for .NET. Compare batch, incremental, and streaming implementations, detect slowdowns against saved results, and evaluate changes intended to reduce processing time and memory allocations.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Comparing batch, incremental, and streaming performance
• Detecting performance regressions after code changes
• Adding benchmarks for a new indicator

Write Python tests with Hypothesis that generate varied inputs and check properties, rules that should always hold. Discover edge cases, reduce failures to simpler examples, and save failing inputs for replay alongside conventional tests.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Finding edge cases in parsers and input handling
• Checking mathematical properties across many inputs
• Testing data encoding and decoding round trips

Turn raw test results into readable reports with pass and fail totals, unstable tests, and coverage, meaning how much code tests exercise. Package results for releases, quality assurance, or reviews, with follow-up actions.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Preparing test reports for a release
• Turning pytest or JUnit output into a readable summary
• Sharing coverage and quality-gate status

Compare current benchmark results, measurements of performance, with historical baselines to identify degradation between builds. Produce a comparison summary, rank affected metrics by severity, and outline questions for further investigation.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Comparing performance between builds
• Reviewing performance degradation in CI/CD results
• Ranking degraded metrics by severity

Record network activity, console errors, screenshots, and page HTML alongside an existing browser automation run. Organize the captured events into searchable files for each page, making it easier to investigate failures and review what happened.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Investigating failed forms or missing page elements
• Adding a trace to an automation run already in progress
• Reviewing network failures and console errors by page

Test how AI agents get started with a product, SDK, or skill by having them attempt real tasks independently. Receive an HTML report with execution records, scores, and concrete suggestions for improving setup and documentation.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Checking whether agents can follow a product's getting-started guide
• Testing whether a SKILL.md delivers its advertised result
• Comparing SDK onboarding across languages and models

Test changed screens or explore an entire app in a real browser. Check interactions, accessibility, and mobile layouts, then document failures with screenshots, reproduction steps, expected results, and suggested fixes.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Checking UI changes in a pull request
• Finding bugs through exploratory testing
• Checking keyboard access and mobile layouts

Build a local browser agent that browses and collects information only from approved domains. Generate a Claude Agent SDK demo with restricted browser actions, then verify blocked external requests through results, audit logs, and a screenshot.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Build a browser agent restricted to approved domains
• Demonstrate blocked navigation to external sites
• Create a scraper with domain restrictions

Check whether completion, bug fix, test, and feature readiness claims are supported by fresh evidence from the current code. Identify missing checks or claims that exceed the evidence, and report whether verification succeeds or requires manual confirmation.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Checking evidence before reporting a task complete
• Confirming the original bug symptom is resolved
• Assessing whether a feature is ready to proceed

Validate a completed feature across all implementation tasks. Checks the full test suite, startup, requirements coverage, component connections, and design boundaries, then reports readiness, concrete failures, or validation that still needs manual attention.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Validate a feature after all tasks are implemented
• Find mismatched data and interfaces between tasks
• Check requirements coverage and design alignment

Measure NVSHMEM communication bandwidth and latency, then package results with GPU placement, hardware connections, exact commands, and raw logs. Produce a portable Markdown report for performance checks and comparisons against a baseline collected under matching conditions.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Running the four-test device communication sanity suite
• Recording same-node and two-node performance results
• Packaging benchmark logs and environment evidence for handoff

Diagnose GPU bugs in CUDA applications with NVIDIA Compute Sanitizer. Select the appropriate tool for memory access errors, shared-memory races, uninitialized reads, or synchronization faults, then build for source locations and interpret the diagnostic report.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Locate illegal or out-of-bounds GPU memory access
• Investigate inconsistent results caused by shared-memory races
• Find reads of uninitialized GPU memory

Benchmark a deployed NVIDIA RAG Blueprint server using a YAML configuration. Measure time spent in each processing stage, run optional aiperf load tests, and compare response latency, throughput, and bottlenecks across concurrency and retrieval settings.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Find slow stages in a running RAG server
• Compare performance across concurrency levels
• Measure the impact of retrieval and reranking settings

Run basic engineering checks for NV-Reason-CXR using chest X-ray images or test fixtures. Test commands with mock responses or run the model locally or through a public API, then inspect JSON results. Not for clinical use.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Check wrapper commands with deterministic mock responses
• Check local dependencies and CUDA availability
• Test live inference through the public API

Turn rough moderation requirements or existing policies into custom safety rules for NVIDIA Nemotron models. Produce a Markdown policy, a JSON category structure, and ready-to-use classification prompts with definitions, exceptions, examples, and rules for keeping conversations on topic.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Turning moderation notes into a structured safety policy
• Adding custom categories and exceptions to an existing policy
• Defining labeling criteria with safe, unsafe, and ambiguous examples