All authors
Claude Skills by sharmapuneet1510
github.com/sharmapuneet1510135 skills0 installs83 views
- DockerContainer configuration - Dockerfile, docker-compose, image optimizationVotes: 0GitHub stars: 3
- FullComplete implementation - code + tests + docs + pipeline in one execution (NO CONTEXT LOSS)Votes: 0GitHub stars: 3
- IacInfrastructure as Code - Terraform, CloudFormation, Kubernetes manifestsVotes: 0GitHub stars: 3
- PipelineCI/CD pipeline generation - GitHub Actions, GitLab CI, JenkinsVotes: 0GitHub stars: 3
- TestComprehensive test generation with 95%+ coverage and business validationVotes: 0GitHub stars: 3
- BuildExecute full-stack generation across all tasks (DB → API → UI → Tests → Deploy)Votes: 0GitHub stars: 3
- ContextBuild project context + architecture + knowledge graphVotes: 0GitHub stars: 3
- IdeateTransform vague ideas into validated project plans with expert feedbackVotes: 0GitHub stars: 3
- NemesisAttack an existing conclusion before it is trusted — compose a specialist persona with NEMESIS, investigate independently, return a verdict with ranked hypotheses of why it failedVotes: 0GitHub stars: 3
- PlanRequirement parsing + task breakdown + strategic planningVotes: 0GitHub stars: 3
- PrPackage deliverables + create GitHub PR + sync artifactsVotes: 0GitHub stars: 3
- ReviewArchitecture review + design validation + gap analysisVotes: 0GitHub stars: 3
- RiskRisk assessment + failure modes + mitigation strategiesVotes: 0GitHub stars: 3
- SolveSolve specific design bottlenecks with multi-dimensional, prescriptive solutionsVotes: 0GitHub stars: 3
- TradeoffComplexity vs. simplicity analysis + 3-option comparisonVotes: 0GitHub stars: 3
- AuditCodebase audit - code health, technical debt, patterns analysisVotes: 0GitHub stars: 3
- Batch ReviewMulti-PR review with unified HTML report and metrics comparisonVotes: 0GitHub stars: 3
- DebugProduction-level root cause analysis with 5-phase RCA workflowVotes: 0GitHub stars: 3
- DiagnoseConversational problem solving for system issues and architectural questionsVotes: 0GitHub stars: 3
- ObserveFour-way conformance comparison — Requirement↔Code, ADR↔Code, Spec↔Code, Tests↔AC. Observations only.Votes: 0GitHub stars: 3
- PerfPerformance optimization with 6-phase bottleneck analysis and roadmapVotes: 0GitHub stars: 3
- QaMaintain the five reusable quality suites — sanity, regression, integration, performance, securityVotes: 0GitHub stars: 3
- ReportUnified quality report - PR review + audit + security + performance synthesisVotes: 0GitHub stars: 3
- ReviewPR validation, code quality scoring, and requirement verificationVotes: 0GitHub stars: 3
- SecuritySecurity audit with 7-phase threat assessment and compliance verificationVotes: 0GitHub stars: 3
- ApproveApprove a spec file or ADR so spec-gate records it — typed by the user after reviewing the fileVotes: 0GitHub stars: 3
- StatusShow which specs and ADRs spec-gate has recorded as approved, and whether gated edits are allowedVotes: 0GitHub stars: 3
- TrivialLet one small change past the spec gate until the user's next message — typed by the user, with a reasonVotes: 0GitHub stars: 3
- Ansible SkillWrite, test, and run Ansible for configuration management: project layout, pinned collections, inventories and variable precedence, roles with argument specs, idempotent tasks with FQCN modules, handlers, validated templates, Vault secrets that never reach logs, check/diff dry runs, rolling updates, and ansible-lint's production profile. The example role here was linted, dry-run, and applied twice (second run changed=0) with ansible-core 2.21.4.Votes: 0GitHub stars: 3
- HTML Run Report SkillThe one shared contract for the HTML report a skill writes at the end of a run that changed something: a single self-contained page outlining what was asked, what was done step by step, every artifact touched, decisions and their labels, gates passed or blocked, and what is left open. Defined once; every other skill's Quick Card names only the sections it adds.Votes: 0GitHub stars: 3
- Jira Implementation SkillTake one Jira ticket from "To Do" to a reviewed, green pull request and back: read the ticket, refuse it without testable acceptance criteria, classify it (new project, schema change, feature, bug, chore) and route it to the skills and gates that class needs, implement test-first with one named test per acceptance criterion, open a PR that carries the criteria matrix, and write the result back to Jira. Every Jira write is shown to the user before it is sent; a human merges.Votes: 0GitHub stars: 3
- Maven SkillSet up, harden, and debug Maven builds: a parent POM that owns every version, a BOM import instead of hand-picked versions, every plugin pinned, Enforcer rules that fail the build on version conflicts, unit and integration tests split across Surefire and Failsafe, a JaCoCo coverage gate, formatting, an SBOM, reproducible jars, and the Maven Wrapper. The example POM here was built with Maven 3.9.16 and JDK 21.Votes: 0GitHub stars: 3
- MSSQL DBA SkillDiagnose and fix a running SQL Server the way an experienced DBA does: find the head blocker and what it is holding, read deadlock graphs, rank waits, find the expensive and plan-unstable queries in Query Store, and review indexes (missing, unused, redundant, fragmented, stale statistics, forwarded heaps) and partitioning (alignment, boundaries, sliding window). Every query here was run against SQL Server 2022 on a lab that reproduced the problem.Votes: 0GitHub stars: 3
- Project Setup SkillBefore the first line of feature code, decide and write down how the project is built: the stack, one library per use case (with the reason and the rejected alternative), project coding rules, the instructions file AI assistants read, the scaffolding (build, lint, format, CI, dependency updates, secret scanning), and an ADR for each major choice — then stop for approval. Library catalog checked against PyPI, npm, and Maven Central on 2026-09-28.Votes: 0GitHub stars: 3
- Python Project SkillSet up and maintain a Python project's build, dependencies, and tooling: uv with a committed lockfile, pyproject.toml as the single config file, src layout, dependency groups, ruff for lint and format, strict mypy, pytest with coverage and warnings-as-errors, pip-audit, a locked CI, a wheel build, and a slim non-root Docker image. Every file here was run with uv, Python 3.14, and Docker on 2026-09-28. Language idioms live in python_advanced_skill.Votes: 0GitHub stars: 3