Skill UI

Install and operate Everything Claude Code (ECC) on the DeepSeek Harness (DSH): native skill roots, first-party hook bridge, bare-insert patch mounting, generator usage, event-support limits, and update workflow. Use when setting up ECC on DSH, wiring hooks into a profile, or debugging missing skills/hooks.

Audit a pytest suite for surviving mutants, untested code, flaky and order-dependent tests, slow tests, and removal candidates, with report-only findings.

This command executes a single experiment iteration for autoresearch workflows. Users review history, decide changes, edit code, commit, and evaluate results. It follows strategy escalation rules to optimize performance systematically without modifying the evaluator.

Audits and governs a team-wide Claude Code rollout: lints shared, local and managed settings, inventories MCP servers and plugins against an allowlist, checks layered CLAUDE.md instructions, and tracks seat adoption with stdlib-only scripts.

Analyzes Claude Code and Codex usage logs to surface cost drivers such as low cache hit rates, unpaid cache writes, top-tier sub-agents, reasoning share, and bloated sessions, with evidence and what-if savings to help cut AI coding spend.

This skill retrieves up-to-date documentation for third-party APIs, SDKs, and libraries using the chub tool. It ensures code accuracy by fetching current references instead of relying on outdated pre-trained knowledge. Use it when users request tasks involving external services like OpenAI or Stripe. Follow chub instructions to search, fetch, and annotate documentation for better future sessions.

Reviews existing test suites for structural problems like brittleness, mock abuse, coverage illusions, slow runs, and poor readability, drawing on twelve classic engineering books such as xUnit Test Patterns and Working Effectively with Legacy Code.