Agent Skills by ZhangSan
Showing 14 vetted skills indexed across 1 GitHub repositories.
kaizen
Record actionable improvements from V-phase retrospectives into .ai_state files.
codex-comm
Bridge Codex communications to CC infrastructure with health checks and session resume.
context7
Retrieve library documentation via CLI and cache findings in notes.
brainstorm
Refine project requirements through Socratic inquiry and produce design options with trade-offs.
reflexion
Enforce a structured self-review checklist before micro-review of software tasks.
code-review
Analyze codebases for T-stage reviews using LLM-as-Judge and output structured JSON.
debugging
Debug software through a four-stage reproduce, locate, fix, and verify workflow.
verification
Run tests, measure coverage, and validate acceptance criteria for software releases.
finish-branch
Automate post-kaizen branch finalization with merge, PR, keep, or discard decisions.
e2e
Automate end-to-end testing workflows for Playwright, Cypress, or Selenium projects.
security-review
Assess API endpoint security, access control, and data handling.
codex-review
Coordinates Codex-based adversarial reviews of project plans across multiple rounds.
plan-first
Convert a confirmed design document into an executable task plan.
claude-delegate
Distribute CC-dispatched tasks to Claude with reverse invocation and verdict signaling.