Jiang tao
Community@sqfzy · 广东广州
A nvimer also a student of Ji'nan University
Agent Skills by Jiang tao
Showing 20 vetted skills indexed across 1 GitHub repositories.
sqfzydesign
Clarify requirements, design architecture through adversarial discussion, and guide implementation with tests and commits.
sqfzyrepeat
Repeat prompts or skill commands on an adaptive schedule until a time limit or condition expires.
sqfzybench
Automate benchmarks and profiling to identify performance bottlenecks across commits.
sqfzytest
Analyze code coverage gaps and generate targeted tests automatically.
sqfzymerge
Merge upstream branches or port features with backup and rollback safeguards.
sqfzyevolve
Assess project maturity and gaps to generate a structured evolution plan.
sqfzyscript
Generate deterministic scripts with error handling and idempotent execution.
sqfzymigrate
Plan and execute incremental software migrations with rollback points and verification.
sqfzyreview
Analyze diffs, PRs, branches, or files to generate severity-organized code review reports.
sqfzyfix
Trace root causes, implement fixes, and validate with regression tests.
sqfzyretro
Generate structured retrospective reports from coding sessions with lessons and root causes.
sqfzydoc
Generates README, API docs, and summaries from scanned source code and files.
sqfzyreport
Generate structured stakeholder reports from git history, artifacts, and code changes.
sqfzyimprove
Orchestrate iterative code quality improvement cycles with adversarial discussions.
sqfzyship
Automate release gating with review, test gap, benchmark, and documentation checks.
sqfzycreate-agent
Generate Claude Code subagent Markdown files with YAML frontmatter.
sqfzydiscuss
Coordinates multi-role discussions and auto-saves results to .artifacts/ on trigger events.
sqfzygit
Generate conventional commit messages and update changelogs from repository state.
sqfzyrefactor
Refactor code structure while preserving behavior by default.
sqfzydebug
Capture error output, classify failure types, and trace root-cause hypotheses into a structured debug report.