Measure output quality, don't vibe it: score a generative task with a two-layer grader — deterministic code metrics + per-dimension LLM-as-judge — over a fixed task set, as signed deltas vs a pinned baseline. Grades cost alongside correctness (pass-slow). Trigger phrases: "eval", "grader", "measure output quality", "LLM-as-judge", "score the output"
Stack-agnostic frontend discipline (web · mobile · desktop): component structure, state, data fetching, loading/empty/error states, i18n, accessibility, performance. frontend-expert-csk applies it on every stack. Trigger phrases: "frontend", "screen", "component", "page", "UI", "state management", "interface"
Refine-to-Done loop: repeat until tests green + review clean + nothing deferred; bounded. Not the harness /loop. Trigger phrases: "iterate", "loop until done", "keep going until"
Stack-agnostic security audit: map the attack surface, trace untrusted input to dangerous calls, surface dependency and configuration flaws. Severity-ranked report with fixes. Trigger phrases: "security scan", "run a security scan", "OWASP check", "scan for vulnerabilities", "find security vulnerabilities", "security audit"
Scope a security audit BEFORE scanning, to cut false positives: map assets, entry points, trust boundaries and 5-8 domain-specific attack classes into a parseable THREAT_MODEL.md. A threat survives a patch; a vulnerability is only evidence for one. Feeds security-scan. Trigger phrases: "threat model", "attack surface", "scope the audit", "trust boundary"
SonarQube quality gate (language-agnostic, local-first): 0 Bugs/Vulns/Hotspots/Code Smells, 0 build warnings. If no analyzer exists, install the language's local server-less Sonar analyzer and run it — never a remote server. Trigger phrases: "sonarqube", "quality gate", "code smell", "sonar scan"
Isolate risky or parallel file-mutating work in a git worktree so the main tree's uncommitted changes are never clobbered. For fan-out agents that edit files, throwaway experiments, or any change you may want to discard cleanly. Trigger phrases: "worktree", "git worktree", "isolate the changes", "sandbox this work", "parallel file edits"
Code review discipline: severity-ranked, reasoned feedback on whether a change improves the system's overall code health. review-agent-csk applies it. Trigger phrases: "code-review", "review the code", "review the PR", "review", "do a review"