Loading...
Loading...
Found 8,529 Skills
Guide for editing speech-led videos where spoken delivery or conversation drives the cut — single-speaker talking-head / 口播, two- or multi-speaker interview / 访谈, video podcast, lecture, tutorial, course, and similar formats. Load before analyzing, planning, or performing non-trivial edits of those formats, including speech cleanup (剪口播 / 口播剪辑 / 去口癖 / clean up fillers / smooth speech), pause or repeated-take removal, motion graphics layered onto the footage (口播加 MG / 加动画), B-roll (加 B-roll / add B-roll), music, or captions. For motion graphics specifically, use this together with the active Motion Graphics skill/workflow available in the current ChatCut environment — this skill adds speech-specific guidance (rhythm-aware timing, frame-aware placement, subject/caption protection, placement verification).
Use when researching a question across multiple sources, comparing evidence-backed options, investigating conflicting claims, or making a recommendation dependent on external evidence. Not for a simple factual lookup, explanation from already-sufficient evidence, code-change review, or diagnosing a concrete software failure.
Assemble actual marketing research and plan sections into a consistent client-ready HTML strategy document and a one-page decision summary.
Perform risk-tiered Code Review of completed code changes: core product logic and high-risk modules receive a five-dimensional deep review in an independent clean context (genuine end-to-end business and packaged-artifact closure, interception of ineffective or side-effecting tests, componentization and functionalization with reuse of existing infrastructure, identification of redundant code and high-maintenance algorithms, runtime-defect scanning for concurrency issues, crashes, and swallowed errors, and security-vulnerability and data/interface compatibility checks), while low-risk UI tweaks receive lightweight verification. Distill recurring rules under “single-source priority, strict admission, and routine pruning.” Use when users ask to review code, review commits or working-tree changes, or identify reinvented implementations and ineffective tests. Dedicated to code review, not reviewing development plans (hand those to oil-plan).
Delegate a clearly scoped, one-off task to Claude Code (cc), WorkBuddy (wb), Antigravity (agy), Codex, OpenCode, DeepSeek Harness (dsh), Grok Build (grok), or Kimi Code (kimi) CLI; return a structured result and resume the original session; handle the initial silent scan, enabled-runner pool, and default selection; and remember which runner to use for each purpose (code review, plan critique, difficult-issue consultation, implementation). Use this skill when the user explicitly requests oil-delegate, unified invocation, or continuation of these external CLI Agents by the full names or aliases above; do not use it for ordinary direct Q&A or an explicit /codex.
Use when making a small, quick change — a bug fix, typo, minor feature, tweak, or anything the user calls "vibe coding" — that looks like 1-3 files in the same module
Do direct-send's unattended work, but stop with commits ready to push to the default branch. Use when the user says direct stop or wants work committed locally without pushing.
Debug Rails issues systematically. Use when encountering ActiveRecord errors like RecordNotFound, routing issues, N+1 query problems detected by Bullet, asset pipeline issues, migration failures, gem conflicts, ActionController errors, CSRF token problems, or any Ruby on Rails application errors requiring diagnosis.
Prevent N+1 queries by eager loading; enable lazy-loading protection in non-production; choose selective fields
Wrap multi-write operations in transactions; use dispatchAfterCommit and idempotency patterns to ensure consistency
Process large datasets efficiently using chunk(), chunkById(), lazy(), and cursor() to reduce memory consumption and improve performance
Create a plan collaboratively with the user, then convert the approved plan into a GitHub issue.