snkrheadz

the-boris-way

Cross-role foundation for everyone: first-principles + honest-reasoning + deep-thinking, life-decision support, teaching/onboarding, rich HTML output, a CLAUDE.md tuner (Boris-way), and a sensitive-file guard hook. Install this first, whatever your role.

PM / business assets maintained by snkrheadz: the 業務定義シート (task definition sheet) as A4 HTML. Pairs with the external phuryn/pm-skills marketplace for the full PM lifecycle.

Engineering workflow pack that fills gaps the official commands don't cover: create-pr, new-skill, prune-redundant-skills, review-inbox, test-and-fix, refactor-swarm, techdebt, trace-dataflow, db-query, plus code-architect, architecture-reviewer, verify-shell, migration, oncall and AWS/GCP advisor agents.

Research pack: arxiv-ai-researcher, gemini-api-researcher, huggingface-spaces-researcher and verify-subagent-result agents for AI/ML paper, API and model discovery plus cross-validation of subagent findings. Role-agnostic — for anyone investigating AI/ML papers, APIs and models.

Personal strategy for the AI-driven shift in work: audit your role's AI exposure (career-ai-scan), map your industry's transformation (industry-ai-map), and find where your expertise x AI opens opportunity (ai-opportunity-scan). Role-agnostic, honest about uncertainty.

Writing pack: strip AI tells from drafts before publishing. stop-ai-slop-jp (Japanese) and stop-ai-slop-en (English) catch predictable AI phrasing, false agency, propositional headings, uniform rhythm, and symbol artifacts. Role-agnostic. Derived from hardikpandya/stop-slop and iKora128/stop-ai-slop-jp (MIT).

Spec-driven development pipeline (the Boris way): scan a repo for agent-readiness, then take a change from a one-line intent to a mergeable PR through human-gated phases — requirement (EARS) → design (DDD/clean-arch) → tasks → implement → review (isolated-context adversarial). Self-contained: bundles the implement phase and the audit rubric. Verification-first, one phase per command.