Miklos Szanyi
Community@swingerman · Odessa, Ukraine
I am a full-stack developer with a passion for designing user-friendly software that effectively addresses your problems.
Agent Skills by Miklos Szanyi
Showing 23 vetted skills indexed across 3 GitHub repositories.
swingermanprototype
Builds rough runnable prototypes from ideas and converts converged ones into specced features.
swingermanpost-merge
Cleans up merged feature branches and resyncs the local main branch after a PR merge.
swingermanzenify-my-terminal
Configure macOS terminal environments with apps, prompts, and plugins.
swingermancrap-analyzer
Rank recently changed functions by CRAP using diff-scoped complexity and coverage signals.
swingermanatdd-mutate
Run mutation testing on ATDD two-stream test suites to measure kill and survive scores.
swingermanatdd
Drives feature development through acceptance test driven development with Gherkin specs and dual test streams.
swingermanatdd-team
Orchestrate a six-phase ATDD workflow with fresh agents and durable handoffs.
swingermanfeature-edit
Update existing feature artifacts and cascade changes downstream after confirmation.
swingermannext
Surveys all DAE project state and recommends the next work item to pick up.
swingermansession-summary
Appends per-session entries to a feature's session-log.md for clean handoffs.
swingermanprogress-log
Propagate DAE handoff summaries into progress.md and tracker records.
swingermanplan
Generates a feature architecture plan with charter validation, review panel, and operator runbook.
swingermanclarify
Resolves ambiguities in a single DAE artifact through an iterative interview.
swingermanonboard
Adopts a project into the DAE methodology by creating its charter, manifest, and consolidation backlog.
swingermanfeature-init
Initialize a DAE feature workspace with folder, metadata, branch, and tracker record.
swingermanprime-context
Load feature context files and ask one follow-up question.
swingermanfix
Drives bug fixes from report through regression spec, mutation gate, and gap analysis.
swingermanreorient
Reloads charter, checkpoint, and task state to restore working context after compaction.
swingermanarch-check
Validates feature code against charter architecture rules for layering, cycles, and forbidden patterns.
swingermanrefine
Reviews and refactors a feature's changed code against charter rules before verification.
swingermandiscover-acs
Discover feature acceptance criteria in domain language and write acs.md.
swingermandiscuss
Explores feature ideas through structured brainstorming and routes them to drop, park, or promote outcomes.
swingermanconsistency-check
Validates DAE artifacts for schema correctness and cross-artifact consistency.
Frequently Asked Questions About Miklos Szanyi
FAQPage SchemaWhat tasks can I accomplish with swingerman's DAE skills?▼
You can onboard projects into the DAE methodology, discover acceptance criteria, draft architecture plans, drive ATDD feature implementation with Gherkin specs, run mutation testing, fix bugs with RED regression specs, refactor green code, and validate artifacts for schema and cross-artifact consistency.
Who are these skills designed for?▼
Full-stack developers and engineering teams practicing acceptance test driven development who want structured, checkpoint-based feature delivery. They suit engineers managing spec-driven projects with charters, trackers, session logs, and architecture fitness rules rather than ad-hoc coding.
How do I use the skills during a typical feature lifecycle?▼
Start with onboard to initialize charter and tracker, use discuss and discover-acs to shape the idea, plan the architecture, implement via atdd with two test streams, then refine, arch-check, and consistency-check before post-merge cleanup. Session-summary and reorient preserve continuity between work sessions.
What prerequisites do the DAE skills require?▼
Skills expect a DAE-onboarded project containing a charter, manifest, tracker, and feature workspace with metadata and branch records. Architecture validation relies on dae_arch.py, ATDD requires a project-specific test pipeline, and post-merge assumes GitHub PR usage via gh pr merge.
Do the skills support recovering context mid-task?▼
Yes. The reorient skill restores role, autonomy boundaries, checkpoint exit criteria, and the current task pointer after context compaction or long runs. The next skill surveys project state to recommend the top action, and prime-context reloads feature files with one follow-up question.