The library

cicd-stage1-example · Simultaneously discovering both failure symptoms (via failure-detector) and contextual causes (via context-analyzer) in CI/CD pipelines to handoff to Stage 2 root-cause agents.

rust-expert · Enforcing senior Rust quality (Decimal vs f64 for money, typed AppError vs Result<_,String>, zero infra imports in domain/) and preventing latent numerical bugs in KoproGo financial computations.

new-command · Creating production Oclif commands for SimpleLogin CLI that follow resource-action patterns, support --format output (plain/json/yaml), and integrate with simplelogin-client SDK with full error handling.

grid-tester · QA verification of glassGRID features (column reorder, row drag, range fill, keyboard nav, virtual scroll) against realistic data before promotion to FEATURES.md—all 36 e2e routes must stay green.

ui-reviewer · Independent post-design QA that verifies rendered output matches design brief, not just code—catches visual regressions, contrast/a11y gaps, and responsive breakages that code review alone misses.

drupal-frontend-dev · Building Drupal theme components with PostCSS styling, proper XSS escaping in Twig, and responsive mobile-first layouts that pass accessibility and CSP checks.

ffi-specialist · Maintaining memory-safe FFI contracts across five languages, where each binding follows opaque-handle + JSON-exchange patterns and all errors surface via last-error thread-local to caller language.

bicep-researcher · Identifying undocumented Bicep features (new decorators, type system, assertions, extensions) that should be added to bicep-docs to keep documentation current with language releases.

mathodology-submission-packager · Final contest submission quality gate—verifies that no secrets/caches exist, all figures are reproducible from source, outside users can run the package, and submission complies with all rules.

system-engineer · Integrating native macOS system capabilities (Spotlight search via mdfind, file metadata via mdls, Finder tags via xattr, Shortcuts automation) into Swift tools via test-driven patterns.

build-validator · Rapidly fixing TypeScript/ESLint/Vite build breakages in a strict React 19 + Express 5 monorepo by identifying the exact root cause (type narrowing, decorator syntax, postinstall failure) and applying minimal fixes.

frank · Production Next.js component implementation from PRDs that faithfully follows design specs, validated in real browser across mobile/tablet/desktop viewports with zero console warnings.

Azure Policy Analyzer · Continuous Azure Policy compliance monitoring and automated remediation generation when new resources violate governance rules.

writer · Producing original long-form thought leadership and explainers on emerging topics when subject-matter expertise is available and citations are verifiable.

gsd-plan-checker · Catching plan incompleteness and hidden assumptions before executing a complex multi-month project that would otherwise fail due to missing coordination or unresolved dependencies.

critic · Independent quality gate after delivery that identifies whether work actually meets the brief (not self-assessed), separates critical fixes from polish, and provides actionable rework guidance.

team-tester · Verifying that frontend, backend, and infrastructure pieces actually work together in realistic scenarios before handing off to production ops.

frontend-engineer · Building production Node.js backends with proper query optimization, transaction handling, authorization checks, and comprehensive error responses that match frontend API contracts.

firmware-engineer · Implementing resource-constrained embedded code with deterministic timing, memory safety, and robust error recovery for hardware control and sensor integration.

add-quiz-questions · Managing multi-language quiz CSV with AI generation and duplicate detection without manual scripts.

security · Auditing code for authorization, RLS, and authentication vulnerabilities before launch.

rule_ars · Validating complex business logic rules against SQL Server data when refactoring allocation algorithms.

cross-page-analyzer · Computing accessibility severity scores and cross-page patterns faster than manual auditing.

ai-eval-engineer · Validating LLM behavior against strict criteria (structure, safety, cost) that string assertions cannot verify.

seo-audit · Identifying crawlability and metadata issues systematically across static Astro sites.

architect · Auditing code for authorization, RLS, and authentication vulnerabilities before launch.

code-inline-reviewer · Validating LLM behavior against strict criteria (structure, safety, cost) that string assertions cannot verify.

build-error-resolver · Auditing code for authorization, RLS, and authentication vulnerabilities before launch.

adr-author · Auditing code for authorization, RLS, and authentication vulnerabilities before launch.

refactorer · Improving code design while guaranteeing behavior preservation through automated validation.

emma-frost · Flagging specific coding-standard violations with inline docs links at scale.

tensorfeed · Auditing code for authorization, RLS, and authentication vulnerabilities before launch.

agentdeals · Handling specialized tasks that generic agents cannot perform.

claude-mem · Handling specialized tasks that generic agents cannot perform.

swarmwage · Handling specialized tasks that generic agents cannot perform.

fal-ai · Handling specialized tasks that generic agents cannot perform.

encode-toolkit · Flagging specific coding-standard violations with inline docs links at scale.

firefox-devtools · Handling specialized tasks that generic agents cannot perform.

iwdp-mcp · Handling specialized tasks that generic agents cannot perform.

pagemap · Browsing high-token pages (PDFs, documentation, dense paywalls) where 97% compression outweighs loss of formatting.

planning-with-files · Multi-session planning where Markdown context outlives any single LLM conversation, across teams using different coding IDEs.

cloudbase-ai-toolkit · Tencent-native development where CloudBase platform provides cost-effective Alibaba-competitive serverless infrastructure for China-market apps.

cloudbase-sites · Two-stage save→deploy workflow for iterative React/Vite front-end development on CloudBase, inspired by Codex Sites pattern.

cloudbase · Tencent stack full-stack apps needing unified auth, NoSQL+SQL, blob storage, serverless, and WeChat Mini Program integration.

owlex · Multi-agent workflows requiring load-balancing or task-specific agents (Codex for code, Gemini for breadth, OpenCode for spec-driven generation).

ejentum · Complex reasoning, code generation, or deception-resistant verification where 679 pre-verified cognitive topologies beat ad-hoc reasoning.

forge · Multi-file refactors, full-stack features, or spec-driven work where parallel validation and retry beat sequential manual oversight.

desktop-commander · Local-only automation across text, code, PDFs, spreadsheets where client-side execution avoids cloud egress and latency.

pty-mcp · Long-lived interactive debugging, DevOps orchestration, or system administration where persistent PTY state beats stateless shell commands.

agenticmail · Multi-agent workflows coordinating via email threads (human-readable audit trail) instead of internal queues or cloud APIs.