Dyad
Official@dyad-sh
Free, local open-source AI app builder
Agent Skills by Dyad
Showing 22 vetted skills indexed across 1 GitHub repositories.
dyad:plan-to-issue
Converts an approved codebase implementation plan into a GitHub issue.
dyad:deflake-e2e-recent-commits
Detects and fixes flaky Playwright E2E tests from recent CI runs and PRs.
dyad:check-workflows
Audits recent GitHub Actions workflow runs and files issues for actionable failures.
dyad:pr-push
Commits changes, runs checks, pushes branches, and creates GitHub pull requests.
dyad:pr-screencast
Records Playwright screenshots demonstrating a PR's feature and posts them as a GitHub PR comment.
dyad:swarm-to-plan
Orchestrates PM, UX, and engineering agents to debate an idea and write a spec to plans/.
dyad:pr-fix:comments
Reads unresolved GitHub PR review comments and addresses or resolves them via the GitHub API.
remember-learnings
Extracts session learnings and updates AGENTS.md and rules files with actionable guidance.
deep-review
Reviews code diffs using parallel finder agents and adversarial verifier agents.
dyad:add-models
Add AI model entries to the language model constants file with researched specifications.
dyad:pr-fix
Resolves GitHub pull request review comments and failing CI checks automatically.
dyad:pr-rebase
Rebases the current branch onto upstream changes, resolves conflicts, and pushes.
dyad:deflake-e2e
Detect and fix flaky Playwright E2E tests by running them repeatedly and investigating failures.
dyad:promote-beta-to-stable
Promotes the latest pre-release to a stable release branch with version bump and push.
dyad:debug-minified-error
Maps minified production error stack traces back to original TypeScript source using source maps.
dyad:debug-with-playwright
Debugs Electron E2E tests by capturing Playwright screenshots at key execution points.
dyad:e2e-rebase
Updates Playwright E2E test snapshots based on failed tests reported in PR comments.
dyad:pr-fix:ci
Diagnoses and fixes failing GitHub Actions CI checks on pull requests from run logs and artifacts.
dyad:fix-issue
Fetches a GitHub issue, plans a fix, and implements it locally.
dyad:feedback-to-issues
Converts customer feedback emails into deduplicated GitHub issues and drafts reply emails.
dyad:multi-pr-review
Orchestrate multiple Claude sub-agents to review PR diffs and post consensus results.
dyad:swarm-pr-review
Orchestrate a three-teammate Claude Code swarm to analyze GitHub PR diffs and post a consensus summary.
Frequently Asked Questions About Dyad
FAQPage SchemaWhat specific tasks does Dyad enable for engineering teams?▼
Dyad enables the automated analysis of GitHub pull request diffs by orchestrating multiple independent review agents. It synthesizes these individual perspectives into a single, consensus-based summary, allowing teams to identify potential issues and architectural improvements without manual intervention during the initial review phase.
Which technical personas benefit most from these capabilities?▼
Senior software engineers, technical leads, and DevOps practitioners benefit most from these capabilities. The platform is designed for teams managing high-velocity repositories who require consistent, multi-faceted code quality checks and accelerated feedback loops during the pull request lifecycle.
What are the core prerequisites for running these review swarms?▼
Users require an active GitHub repository and valid credentials for the underlying Claude models. The system functions by interfacing directly with GitHub PR diffs, necessitating appropriate repository permissions to post consensus summaries and review comments back to the platform.