石田 翔大
Community@hitsumabushi334 · Aichi prefecture
Shota Ishida I` m studying at Nagoya Institute of technology. I am majoring in bioinorganic chemistry and currently studying Python and JavaScript.
Agent Skills by 石田 翔大
Showing 17 vetted skills indexed across 1 GitHub repositories.
using-git-worktrees
A tool for creating and managing isolated Git worktrees efficiently and safely, ensuring your main workspace stays clean and your builds are reproducible. It .----This is the wrong one, ignore it. Let's try again.
test-driven-development
Enforces Red-Green-Refactor cycles with failing tests before code for automated test validation.
systematic-debugging
Guide engineers through a four-phase root cause debugging process.
prompt-engineering
Improves LLM prompt reliability through few-shot learning, chain-of-thought prompting, and template systems for agents and sub-agents.
using-superpowers
Invoke the Skill tool before responding to load applicable skills.
dispatching-parallel-agents
Dispatch independent tasks across parallel AI agents for concurrent investigation.
context-engineering
Structure agent context with progressive disclosure and attention budgeting.
executing-plans
Execute implementation plans in review-controlled batches with progress reporting.
finishing-a-development-branch
Finalize a development branch with four merge or cleanup options after tests pass.
brainstorming
Guide structured ideation through one-question-at-a-time dialogue.
writing-plans
Generate detailed implementation plans from specifications for software tasks.
requesting-code-review
Dispatches a code-review agent to guide severity-based feedback on tasks and merges.
receiving-code-review
Evaluate code review feedback through verification and incremental testing.
serena-development
Coordinates coding tasks through Serena MCP for exploration, editing, and verification in LSP-enabled projects.
writing-skills
Apply RED-GREEN-REFACTOR workflows to test and validate Skills before deployment.
verification-before-completion
Run verification commands and report completion only with fresh evidence.
subagent-driven-development
Dispatch independent subagents for each task with two-stage reviews.