استكشف منشورات X المراجَعة عن المهارات، ومن يشاركونها، والمشاريع المذكورة فيها.
يشمل العدد فقط منشورات المهارات التي جُمعت وروجعت. اليوم الفارغ يعني عدم جمع منشورات هنا. التواريخ بحسب UTC.
240 منشورات جُمعت
The creator releases the HeyGen video Skill used in their Codex workflow on Mac. It generates short clips with sound from text, images or reference media; the quoted article explains the surrounding workflow.
Luobo PPT is an open-source Skill for turning articles, scripts or outlines into editable presentations with speaker notes, or restyling an existing deck from a reference image while checking its original text. The author shows style variations, a mathematics lesson and an article-to-slides example; illustrations can be drawn programmatically without an image-generation model.
Jinghong B-roll plans supplementary shots from a narration brief, submits generation jobs through AutoDL Art and downloads named media files. The creator documents MiniMax H3 reference modes and IndexTTS2 voice generation; final editing remains in the user’s editor.
The creator shares genvid-collage, a Skill for documentary-style product videos with generated collage visuals, timed narration and Remotion rendering. The workflow requires configured image, video and speech tools.
landable helps find open-source repositories that review outside contributions, then develops a scoped issue into a pull request. The creator emphasizes checking contribution opportunities before coding, while the human chooses the work and approves publication.
QuantumTransf shares a Skill that audits and rewrites bloated READMEs using Art of README principles so readers can judge fit quickly instead of being sold to.
Bart shares his motion-graphics repository and evolving video workflow, from rough cuts and foreground separation to animation overlays and export, with a work-in-progress short-video example.
Bole Chen open-sources serverless-cost-guardrails to self-check serverless setups for paths that can produce surprise runaway bills, and invites trials and PRs.
qiaomu-highlight-to-social-content turns the user's Qiaomu Clipper highlights and notes into sourced subtitle images and social copy. Its creator presents it as an independent layer above Qiaomu Clipper, with confirmed quotes, real video frames and checked background information.
This article explains why the creator built a HeyGen video Skill: refine a brief, choose settings, review the estimated cost, then generate and download a clip after confirmation. Prompt refinement can be used separately from paid generation.
imknight open-sources a HeyGen video Skill used on Codex Mac to generate videos from text/image inputs with sound.
Bhavesh Chowdhury shares a Skill that helps AI agents build, adapt, or review React and Next.js charts with @kind-ui/charts—covering chart selection, composition, styling, legends, loading, and accessibility.
majiang-photo-story turns a real photo into a white-background poster with sparse black-line characters and a story drawn from the subject. In a quote of her own release post, the creator explains preserving the photographed object, extending the scene into the whitespace, and checking caption wording separately from lettering.
pi-matt-implement-flow adapts a Matt-style ticket graph into an implementation workflow for pi, with a ledger and audit report. The creator shares a run report and explains assigning coding, review and integration roles around the dependency frontier.
Dimabytes announces herdr-fleet's first merged PR: a Skill that lets Claude Code drive Codex, Devin, Cursor, and Command Code in parallel through Herdr panes, and invites others to try and contribute.
Jared Palmer quotes a write-up on fine-tuning Kev-4B and points to the official kev-finetune Skill, which finds decision-model callsites, helps generate synthetic data, walks through fine-tuning on Modal, and pairs with kev-deploy.
karaoke-subtitles burns word-synced English captions into video, highlighting each spoken word and optionally adding short Chinese glosses. The creator presents it as a video preparation tool for English learners.
In a discussion about AI video editing, the creator shares design-engineering. Its source includes guidance for product launch video sound and multi-scene cuts alongside interface design and motion.
shibu0x built parrot so coding agents learn his local writing voice from past agent messages and social posts (with secrets redacted), then draft or rewrite tweets/posts/emails like him and flag AI stock phrases; this tweet was written with it.
trigger-doctor diagnoses why an existing Skill description misses relevant requests or catches unrelated ones. The creator presents a workflow combining structural checks, simulated trigger judgments, a proposed description and a saved regression suite.