debugger
25Agent
Perform systematic root-cause analysis and apply minimal fixes across the project stack.
4 7d ago A 18 tokens
Agent
Find all code affected by changes (callers, tests, imports) in Phase 2 (read-only) or Phase 6 (edit-capable).
4 7d ago A 36 tokens
Agent
Investor-mode reality check on the project's current state — audits all sub-apps/products (read from project.json → apps and CLAUDE.md) and the underlying business thesis with VC-style critique. Read-only.
4 7d ago A 48 tokens
Agent
Coordinate feature implementation across database, API, UI, deploy, and promotion layers.
4 7d ago A 20 tokens
Agent
Validate a proposed action against risk rules and emit APPROVED/REJECTED with risk level and rollback plan, invoked cross-phase before risky operations.
4 7d ago A 34 tokens
Agent
Execute Phase 4 code changes as a leaf executor (stepfile input), or orchestrate step-by-step execution of a ready workspace plan (taskdir input) with per-step verification loops.
4 7d ago A 41 tokens
Agent
Break down large tasks (>1 week) into multi-phase plans with phase docs and copy-paste agent prompts.
4 7d ago A 26 tokens
Agent
Database migration safety — Prisma SQL and Prisma schema validation.
4 7d ago A 14 tokens
Agent
Execute safe, incremental code and schema migrations with rollback at every step.
4 7d ago A 17 tokens
Agent
Analyse agent and application performance metrics and produce actionable optimization reports.
4 7d ago A 15 tokens
Agent
Measure, rank, and fix performance bottlenecks with human approval before each change.
4 7d ago A 21 tokens
Agent
Compare implementation against original plan; emit APPROVED/NEEDS CHANGES/REJECTED verdict.
4 7d ago A 24 tokens
Agent
Detect task completion and delegate structured context payload to @task-documenter for documentation.
4 7d ago A 21 tokens
pr-generator
38Agent
Generate PR titles, descriptions, and review checklists from changes.
4 7d ago A 16 tokens
Agent
Generate structured prompts for executor agents containing objective, file paths, code examples, and falsifiable acceptance criteria.
4 7d ago A 25 tokens
Agent
LLM-as-Judge quality evaluation + non-blocking browser smoke in Phase 5; report-only. Deterministic build/typecheck/lint/test verdicts are @verification-agent's job — this agent's deterministic re-runs are input to the LLM score, not the gate.
4 7d ago A 62 tokens
Agent
Build and refactor React 19 / Next.js 15 components with performance budgets and accessibility gates.
4 7d ago A 24 tokens
Agent
Produce Phase 9 retrospective artifact with lessons learned, bias checks, metrics, and improvement recommendations.
4 7d ago A 24 tokens
Agent
Perform OWASP Top 10 checks and domain-specific STRIDE threat modeling (project.json → domainTerms.threatModelExample) when orchestrator needs security validation in Phase 5 Quality Gate.
4 7d ago A 43 tokens
Agent
Audit code for silent failures including empty catches, swallowed errors, and missing error propagation.
4 7d ago A 23 tokens
Agent
Orchestrate end-of-sprint read-only audit by scoping recent changes, fanning out audit subagents, running check-only suites, and producing a PASS/FAIL report.
4 7d ago A 40 tokens
Agent
Orchestrate SRE tasks (SLO definition, incident response, capacity planning, toil reduction) with human gates on all destructive operations.
4 7d ago A 34 tokens
Agent
Produce the canonical SUMMARY.md final report (Phase 8) for completed tasks with quantified metrics, cited data sources, and role-specific content.
4 7d ago A 31 tokens
Agent
Read a retrospective digest of the whole agent system and write an evidence-cited analysis of what hurts, why, and what to change across agents, skills, commands, hooks and processes.
4 7d ago A 42 tokens