itzTiru
UserSkills I use with Claude Code every day
Categories
Indexed Skills (5)
cost-lean-orchestration
Token-efficient model routing for complex engineering sessions. The main model (Opus / Fable, high reasoning) never bulk-reads or explores — it delegates all exploration, bulk file reading, codebase search, log scanning, dependency tracing, and documentation reading to Sonnet subagents that return compact structured findings, and it spends its own tokens only on synthesis, architecture, judgment, and final code. Use this skill in EVERY session that involves reading more than a couple of files, exploring or mapping a codebase, investigating a bug, auditing call sites, reviewing a large diff, digesting logs or docs, or any multi-step engineering task — even if the user never mentions cost, tokens, subagents, or delegation. This is a standing operating discipline, not an on-request feature.
fable-book
Use when asked to turn a PDF, Markdown, HTML, or text source into a beautiful iPad-readable book or "fable book", to restyle a document in the Anthropic/Fable theme (warm ivory, coral, serif), or to make a long technical document "extremely readable" as a PDF for tablet reading.
human-article-writing
Write, rewrite, or edit articles, LinkedIn posts, blog posts, essays, newsletters, and other long-form prose so the result reads as high-quality, natural human writing with none of the stylistic patterns typical of AI-generated text. Use this skill whenever the user asks for an article, post, essay, opinion piece, or long-form draft, whenever they ask to "humanize", "rewrite", "polish", or "fix the tone" of existing text, and whenever they mention AI-sounding writing, detection, or wanting text that reads like a person wrote it — even if they don't say "article" explicitly.
prompt-alignment-audit
Use when reviewing, auditing, or optimizing any LLM prompt (system prompt, extraction prompt, template) against provider guidance — when a prompt underperforms, after a model upgrade or version switch, before shipping a new production prompt, or when asked whether a prompt follows current Anthropic/OpenAI/Gemini best practices. Also use when asked to "check my prompt", "improve this prompt", or fetch the latest prompting guidance (--refresh).
shipit-lite
Use before pushing commits to any remote, when asked "is this safe to push", "check before push", or "shipit" — and whenever the pre-push hook flags findings. Catches scratch/junk files, secret-looking content, and unwanted AI co-author trailers in unpushed commits, and guides the fix for each.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.