en un clic
cloud-native-ref
cloud-native-ref contient 10 skills collectées depuis Smana, avec une couverture métier par dépôt et des pages de détail sur le site.
Skills dans ce dépôt
Validate a spec directory end-to-end — structural completeness, FR/SC counts, falsifiable success criteria, FR coverage in plan tasks, no stale CL-N references, and constitution compliance. Single gate before implementation.
Create a specification for non-trivial platform changes. Creates a GitHub issue anchored to a spec directory with the SDD template. Type is auto-inferred from the description; pass an explicit type to override.
Resolve [NEEDS CLARIFICATION] markers — present 2–3 structured options, append the chosen decision as a CL-N entry to clarifications.md, replace the marker in spec.md with the CL-N reference. Append-only — never overwrites prior deliberations.
Create or update a Pull Request with AI-generated description, mermaid diagram, file walkthrough, and automatic SDD spec detection. Uses templates/pr-body.md for structure.
Verify that a merged spec's success criteria (SC-XXX) are actually met in the live cluster. Deploys the example manifest, watches Flux reconciliation, queries VictoriaMetrics/VictoriaLogs for evidence, writes docs/specs/<dir>/VERIFICATION.md.
Research patterns, ecosystem tools, and best practices before filling a spec. Runs in a forked Explore subagent so it can query Context7, WebSearch, and the whole codebase without consuming the main context window. Writes docs/specs/NNN-slug/research.md.
Show a pipeline overview of all specifications — counts of Draft / Implementing / Done, plus stale-spec detection.
Validates Crossplane compositions end-to-end — KCL formatting, syntax, rendering, and security/policy checks (Polaris, kube-linter, Datree). Auto-activates for KCL and composition files.
Run pre-commit validation before committing. Use for automated hooks, Terraform cleanup, or emoji reference. For simple commits, Claude's native capability suffices.
Comprehensive PR analysis with security review, code quality assessment, and automatic improvements. Use when reviewing PRs for security issues, performance optimizations, or best practices violations. Provides multi-dimensional analysis with safe auto-apply options.