Desenvolvimento

Self-referential loop until task completion with configurable verification reviewer

Use when implementing any feature or bugfix, before writing implementation code

Goal-driven smoke test for the feature just implemented. Drives a browser (UI) or hits the running service (backend) — both, when full-stack. Tests every behaviour acceptance criteria define plus exploratory edges, catches console errors / network failures / a11y issues / contract violations / auth leaks / missing observability, fixes every gap or bug it finds, then re-runs until clean. Captures t

Audit healthcare and caregiving software for provider burnout risk factors. Analyzes workload distribution fairness, scheduling equity, documentation burden, alert fatigue indicators, break and rest compliance, overtime patterns, and systemic contributors to staff burnout. Produces a burnout risk scorecard with actionable recommendations tied to patient safety outcomes. Use when you need to audit

Autonomously review and improve existing code through up to 5 iterations of analysis, domain verification, fixing, and validation.

Process-first advisor routing with artifact capture

Comprehensive regression model diagnostics and assumption checking. Use when validating regression models, checking assumptions (linearity, homoskedasticity, normality, independence), detecting outliers/influence, testing multicollinearity, or when user mentions residuals, heteroskedasticity, VIF, Cook's distance, or diagnostic plots.

Use when deciding where code should live, organising files, or auditing project structure. Checks colocation, grouping, and directory anti-patterns.

Audit commercial lease compliance systems -- CAM reconciliation accuracy (pro-rata share, caps, admin fees), lease abstraction completeness, critical date tracking and deadline alerting, clause compliance monitoring (insurance certificates, permitted use, co-tenancy), tenant obligation enforcement, and ASC 842/IFRS 16 lease accounting (ROU assets, lease liabilities, discount rates). Use when revie

Self-improving skill system. Reads /recall and /metrics output, identifies which skills need patching based on rework patterns and regression data, generates additive patches, and logs changes. Evolves skills based on learnings from any tech stack.

Methodically walks a running app — every public route, every interactive control, every state (empty/loading/error/auth/un-auth) — to find real bugs, then root-causes, fixes, adds a regression test, and verifies each one before moving to the next. Use when: 'find bugs in the app', 'sweep for regressions', 'methodically hunt and fix bugs', 'audit user flows', 'find broken buttons / forms / links'.

Use when designing, running, debugging, or hardening deterministic eval suites for agent skills, prompts, tool workflows, or MCP-backed cases.