Spawns a GPT-5.4-high agent for planning, review, analysis, and debugging of complex or difficult tasks—higher-quality reasoning, second opinions, or architectural guidance when ambiguity, risk, or failures appear. ONLY call this subagent when the user explicitly requests it. NEVER use the fast model for this…
Adversarial review plus acceptance verification. Scrutinizes diffs for quality issues, runs tests and lint when possible, and reports review findings alongside pass/fail against completion criteria. Use after substantive edits when you need review plus verification; use test-runner for execution-only test runs.
Creative design specialist for visual direction, composition, interaction concepts, and design exploration. Use when shaping the look and feel before implementation.
Technical UI implementation specialist for high-fidelity interfaces. Use when translating approved designs, mockups, or Figma files into polished, design-system-compliant frontend code.
Technical devil's advocate that challenges feature plans, strategies, and implementation approaches when the user launches a request in plan mode. This subagent should be used by the Ask Clarification Tool before creating a plan when planning features, designing systems, or proposing solutions. It should be triggered…
Documentation specialist for long-form developer docs, API references, guides, and substantial README updates. Use when documentation itself is the deliverable.
Writes and extends unit tests and end-to-end tests that match the project's frameworks and conventions. Use when implementing features, fixing bugs, or when coverage is required for new or changed code.
Test execution specialist that proactively runs tests and reports results. Use when you need execution-only verification; use code-skeptic when you also want adversarial review.
★not rated 39 4mo agoA38 tokens
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: