mmdmcy
Community@mmdmcy · The Netherlands
☦️ Christian first, AI & ML Engineer second.
Agent Skills by mmdmcy
Showing 44 vetted skills indexed across 1 GitHub repositories.
mmdmcyfigure-it-out
Designs an auditable multi-phase workflow for large migrations and cross-cutting changes.
mmdmcybro
Restates the previous AI message in plain language without jargon.
mmdmcyinterrogate
Spawns multiple LLM reviewers to adversarially review code changes and synthesize a verdict.
mmdmcyprinciple-boundary-discipline
Concentrate validation and error handling at system boundaries while keeping business logic in pure functions.
mmdmcysetup-pstack
Configure per-role model assignments for pstack by writing a role map file.
mmdmcyswarm
Orchestrates parallel cloud workers to cover slices or race briefs and returns one aggregated report.
mmdmcyprinciple-model-the-domain
Guides encoding domain logic into data structures instead of scattered conditionals.
mmdmcytechnical-writing
Applies Diátaxis, Google style, STE, and Global English rules to technical documentation.
mmdmcyprinciple-fix-root-causes
Guides debugging by tracing symptoms to root causes instead of applying symptom-level workarounds.
mmdmcywhy
Investigate design rationale behind code by querying seven evidence sources in parallel.
mmdmcyprinciple-encode-lessons-in-structure
Converts recurring corrections into lint rules, metadata flags, runtime checks, or scripts.
mmdmcyhow
Explains codebase architecture and subsystem flows through parallel exploration agents.
mmdmcyblast-radius
Analyzes what a code change could break beyond the diff and proves safety facts by running real code.
mmdmcyprinciple-guard-the-context-window
Route large payloads to subagents and keep summaries in the main context window.
mmdmcyprinciple-subtract-before-you-add
Guides removal of dead code and redundant validators before adding new features.
mmdmcyreflect
Spawns parallel review subagents over session transcripts to route learnings into skill edits.
mmdmcyprinciple-foundational-thinking
Guides structural decisions on data structures, scaffolding order, and concurrency before writing logic.
mmdmcyshow-me-your-work
Maintains an append-only TSV decision log recording choices, evidence, and results for reviewable long-running work.
mmdmcyprinciple-build-the-lever
Guides building rerunnable codemods, scripts, and generators instead of manual edits.
mmdmcypoteto-mode
Orchestrates multi-step coding workflows with playbooks, subagents, and verification gates.
mmdmcyprinciple-make-operations-idempotent
Designs state-mutating operations that converge to the same end state after crashes and retries.
mmdmcyrecall
Reconstructs recent working context from chat transcripts and shared project records into a status brief.
mmdmcyprinciple-sequence-verifiable-units
Sequences multi-step work into small units verified before advancing.
mmdmcymaintain-verification-skill
Audits and corrects a project's verification skill and feature map through source review and live app drives.
Frequently Asked Questions About mmdmcy
FAQPage SchemaWhat tasks can I accomplish with mmdmcy's skills?▼
You can orchestrate parallel agent workers (swarm, arena), run adversarial multi-model code reviews (interrogate), analyze change blast radius, debug to root causes, reconstruct prior work context (recall), explain architecture (how, why, teach), and enforce 20+ coding principles like idempotency, type-system discipline, and boundary validation.
Who are these skills designed for?▼
Software engineers and AI/ML practitioners working with agentic coding assistants. The skills target developers doing migrations, refactors, code review, debugging, and documentation who want structured playbooks, parallel subagent orchestration, and verifiable execution rather than ad-hoc prompting.
How do the orchestration skills like swarm and arena work in practice?▼
Invoke them by name (e.g., $swarm or $arena) inside an agentic coding environment. Swarm fans out N parallel workers and drains results into one report; arena spawns competing candidates for the same task, then grafts the strongest parts of losers into a chosen base.
What prerequisites or dependencies do these skills require?▼
Skills assume an agent runtime supporting subagents and skill invocation (e.g., $skill-name syntax). The why skill discovers available MCPs for source control, issue trackers, and observability; setup-pstack configures per-role model maps; create-verification-skill needs a runnable app to drive.
Do these skills support TypeScript projects specifically?▼
Yes. The typescript-best-practices skill applies automatically when reading or editing any .ts or .tsx file, and principle-type-system-discipline enforces making illegal states unrepresentable, branded primitives, and parsing external data at boundaries in any statically-typed language.