MyACE makes your AI coding agent's rules, skills, and workflows portable.
Agent
MyACE project-specific orchestrator. Extends the global orchestrator with MyACE-specific subagents, commands, and conventions for the three-component architecture (backend FastAPI, frontend React/Vite, CLI Typer).
Agent
Builds and maintains LLM-integrated features — RAG pipelines, tool-calling integrations, structured-output parsing, embedding/vector search — treating model output as untrusted input requiring validation.
Agent
Curates what goes into an agent's context window across a long-running or multi-agent task — deciding what stays, what gets summarized, and what gets dropped.
Agent
Designs and iterates on prompts, system instructions, and agent/skill copy — owns defining what "better" means for a change before making it, not after.
Agent
Builds Jetpack Compose UI, ViewModels, Room database, WorkManager tasks, and Firebase integration.
Agent
Read-only reviewer for Jetpack Compose recomposition, ViewModel scope, permission handling, and Play Store guidelines.
Agent
Read-only agent that reviews changes against whichever compliance or governance framework the project declares, flagging gaps like missing threat models or missing data-classification notes rather than assuming a specific framework.
Agent
Read-only agent that reviews diffs and code for OWASP-Top-10-style vulnerabilities, auth/authz gaps, and secrets, producing a structured findings report instead of fixing anything itself.
Agent
Read-only reviewer focused on API consistency, migration safety, and boundary-validation gaps — flags problems precisely, never edits code itself.
Agent
Implements API and service changes end to end — schema, migration, endpoint logic, and tests land together, with boundary validation and idempotency treated as part of the feature, not an afterthought.
Agent
Builds CI/CD pipelines, container images, deployment configs, and observability dashboards.
Agent
Read-only reviewer for pipeline safety, artifact integrity, secret handling, and deployment blast radius.
Agent
Reviews existing documentation against the code/config it describes, flags drift and unclear passages, and proposes specific fixes without silently rewriting meaning.
Agent
Writes new documentation for a given change, matching the project's existing structure, tone, and terminology, with edit access scoped to documentation files.
Agent
Read-only agent that classifies an AI system into the EU AI Act risk tier (unacceptable/high/limited/minimal) and identifies the actor role (provider/deployer/importer/distributor), using the risk-classification skill's decision tree.
Agent
Read-only agent that reviews code and design against the applicable EU AI Act obligations (Arts 8-15, 26, 50, 53-55), producing PASS/FAIL/N/A findings with article citations.
Agent
Read-only agent that drafts the EU AI Act compliance documentation set — technical documentation (Annex IV), EU declaration of conformity, risk-management file, data-governance records, and GPAI training-data summary — as structured drafts for human/legal review.
Agent
Implements UI changes end to end — components, styling, state — and always finishes with a real visual-verification pass before calling the work done.
Agent
Read-only review of a UI diff for accessibility, visual consistency, and responsive-behavior issues — flags problems, never edits.
Agent
Read-only reviewer for schema design, migration safety, query performance, and data-integrity gaps — flags problems precisely, never edits code.
Agent
Implements database schema, migration, and query changes — with constraints, indexes, and rollback paths treated as part of the change, not an afterthought.
Agent
Implements features end to end across the stack — backend endpoint, frontend component, integration wiring, and e2e test.
Agent
Read-only reviewer that checks frontend-backend integration — contract adherence, real HTTP path, error propagation, UI states.
Agent
Writes and validates infrastructure-as-code changes — scaffolds resources, runs plan/lint/policy checks — but never applies them to real infrastructure without explicit human approval.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: