constellation

Adversarial code reviewer. Assumes the worst about every change. Identifies failure modes, security gaps, race conditions, and edge cases that optimistic reviewers miss. Read-only — critiques but never fixes.

Use this agent when you need a thorough code review that balances engineering excellence with development velocity. Invoke after completing a logical chunk of code, implementing a feature, or before merging a pull request. Focuses on substantive issues while also addressing style.

Use this agent for security-focused code review to identify HIGH-CONFIDENCE vulnerabilities with real exploitation potential. Minimizes false positives with >80% confidence threshold. Based on Anthropic's claude-code-security-review methodology.

Read-only root-cause investigation agent. Traces a bug to its source through the four-phase discipline, runs the failing case to observe behavior, and returns a structured root-cause report. Cannot edit — it investigates and recommends, the controller applies the fix. Invoke inside Workflow scripts to isolate one bug per agent, or for any large investigation that would pollute the controller's context.

Strict TDD enforcement agent. Guides the RED-GREEN-REFACTOR cycle, validates test-first compliance, and blocks implementation without failing tests.

Use when creating, editing, reviewing, or rendering software architecture or flow diagrams — C4 diagrams (system context, container, component, deployment, dynamic), sequence diagrams, request-flow visuals, or any Mermaid authoring where the rendered output must be readable. Also use when a Mermaid diagram renders with overlapping labels, parse errors, or a single-column layout, or when the user asks to "diagram the architecture", "draw the request flow", "map the sequences", or wants LucidChart-quality output from Mermaid.

Use when about to create a feature, build a component, add functionality, modify behavior, or start any new work — or when the user asks how should we, what approach, which way, explore options, design X, or spec this out. Triggers before any code, scaffold, or implementation skill runs.

Use when you want brutal, worst-case review of code resilience -- failure modes, race conditions, blast radius, and edge cases optimistic reviewers miss.

Use when reviewing a pull request or diff, giving PR feedback, validating an implementation against a Jira ticket or spec, after completing a feature or task, before merging to main, or whenever you need a fresh set of eyes on changes before they ship

Use when committing, opening a PR, creating or naming a branch, resolving a merge conflict, setting up a worktree, squashing, rebasing, or wrapping up a finished branch — i.e. any hands-on git operation where the output (commit message, PR body, branch name, conflict resolution) needs to be correct.

8 hooks across 5 events