mit einem Klick
agent-runtime-kit
agent-runtime-kit enthält 20 gesammelte Skills von shir-bruchim, mit Repository-Berufsabdeckung und Skill-Detailseiten auf SkillsMP.
Skills in diesem Repository
Promotes recurring feedback into the right skill, then guides /compact at phase boundaries.
Testing guidance for pytest, Jest/Vitest, Go, and TDD. Use when writing tests or improving coverage.
Methodical debugging with evidence and hypothesis testing. Use when troubleshooting fails or root cause is unclear.
Create new skills, commands, hooks, or subagents. Use when adding capabilities to Claude Code or Cursor.
PostgreSQL patterns for queries, schema, indexing, security. Use when writing SQL, designing schema, or adding indexes.
Reviews a GitHub PR diff for correctness, security, tests, architecture. Use when asked to review a PR or pull request.
Orchestrates the Ralph pipeline (spec-interview → PRD → execute). Use for features needing autonomous implementation.
Verification driven by built-in /loop. Use when verifying an implementation or running a final check before commit.
Write design docs, READMEs, Jira tickets, RFCs — with or without diagrams. Diagrams render via draw.io MCP plugin.
REST API design patterns. Use when designing endpoints, defining schemas, or establishing API conventions.
CI/CD, health checks, and rollback patterns. Use when deploying, setting up CI/CD, or writing GitHub Actions.
Docker + Compose patterns for dev, security, networking, volumes. Use when writing a Dockerfile or docker-compose.
Generates PRDs through guided discovery. Use when starting a feature or defining requirements.
Git operations for commit, push, PR, and ship workflows with safety rules. Use for any version control task.
Implements Jira tickets end-to-end (git + plan + code + test). Use when given a Jira ticket ID (e.g., PBAT-123).
Create phase-based plans with file paths and verification. Use for features, refactors, or multi-step work.
Converts PRDs into prd.json for Ralph. Use when preparing PRDs for Ralph or breaking features into atomic stories.
Security review and safe-op patterns. Use when reviewing for vulnerabilities, setting up hooks, or protecting files.
Interviews users to build project specs through guided discovery. Use when starting a new project or feature.
Deep web research via WebSearch + WebFetch. Modes: search, reason, research. Use for facts or cited reports.