David Sifry
Community@dsifry · San Francisco, CA
Agent Skills by David Sifry
Showing 31 vetted skills indexed across 4 GitHub repositories.
create-issue
Create structured GitHub issues with embedded agent instructions for PR lifecycle management.
plan-review-gate
Spawn three independent reviewers to assess implementation plan feasibility, completeness, and scope alignment.
orchestrated-execution
Orchestrate a 4-phase execution loop with implementation, validation, adversarial review, and commit.
external-tools
Delegate coding and review tasks to external AI CLI tools.
visual-review
Capture web page and UI screenshots with Playwright for visual inspection.
Create GitHub Issue Skill
Create GitHub issues with problem statements, technical specs, and TDD plans.
beads-orchestration
Coordinate AI agents to manage BEADS epic and GitHub issue lifecycles.
brainstorming-extension
Trigger a five-agent design-review gate after design documents are committed.
design-review-gate
Coordinate five domain experts to approve design documents through parallel review.
using-git-worktrees
Create isolated Git worktrees for feature development with automated setup.
test-driven-development
Enforce a test-first workflow using the Red-Green-Refactor cycle.
testing-anti-patterns
Detect testing anti-patterns like improper mocking and incomplete mocks in codebases.
systematic-debugging
Enforce a four-phase debugging workflow with hypothesis testing before fixes.
sharing-skills
Create branches, edit skills, and submit PRs to upstream repositories.
using-superpowers
Enforce a mandatory skill discovery and TodoWrite checklist workflow.
dispatching-parallel-agents
Identify independent failures and dispatch parallel Claude agents per domain.
executing-plans
Executes staged development plans with checkpoints and progress reporting between batches.
finishing-a-development-branch
Evaluate test results and offer merge, PR, keep, or discard options.
root-cause-tracing
Trace errors backward through call stacks to identify root causes.
condition-based-waiting
Replace arbitrary timeouts with condition polling in test automation.
brainstorming
Guide teams through incremental questioning to refine ideas into structured designs.
testing-skills-with-subagents
Run RED-GREEN-REFACTOR testing cycles on documentation-based skills.
writing-plans
Generate implementation plans with file paths, code examples, and verification steps.
requesting-code-review
Dispatch the code-reviewer subagent to validate implementations against the plan.