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.
51 منشورات جُمعت
Autohand releases four Skill packs for code review, Next.js App Router, Rails and Django. Each pack provides task-specific playbooks rather than a single representative Skill.
The maintainer introduces Research Agent Skills, a collection of playbooks for literature reviews, citation checking, scientific writing and research workflows across disciplines.
dontbesilent updates the dbs router to read each Skill’s handoff metadata and coordinate multiple Skills for one task.
Abdullah Formuli shares Skill Audit, which profiles agent Skills, flags overlaps and compares competing Skills side by side.
molang shares chinese-natural-reply, a Skill for concise, direct Chinese user replies; the repository includes before-and-after examples.
Bryon Larrance releases SOC, a Skill that interviews the user before writing accepted agent harness rules into a repository’s .agents directory.
Baur Rustem shows how compact-receipts reports what a Claude Code context compaction dropped, including environments and files; an optional hook restores that context immediately after compaction.
The author shares desktop-handoff for moving a conversation into a new Claude Code session in the Claude desktop sidebar. The post gives a macOS-only install and trigger, with tested app and CLI versions.
Max Hou introduces CLI Doctor, a read-only environment check for Homebrew, npm, pip, Rust, Bun, and uv that reports upgrade risks and proposed commands without running them.
OpenRig creator Michael Schwarz shares its rigs Skill for installing and coordinating an agent team from local Claude Code or Codex sessions.
Max Hou shares CLI Doctor, a read-only Skill that checks installed npm, Homebrew and pip command-line tools for major version jumps and dependency impact before an upgrade.
Explains dzhng’s feature workflow: explore unknowns, write a specification, run an implementation loop and review the agent’s recorded decisions. The multi-day run cited in the post is the project author’s example.
Matteo releases Skills for crypto due diligence, interview preparation, and whiteboard explainer videos, plus beer-can labels and 3D-printable parts from side projects. Each starts by asking what the user needs.
Orbital Desk built Honeypot as a free, machine-readable index of AI agent Skills. Agents can read one JSON file to find entries and their installation instructions, without signing up.
Sahil Gulihar asks users to try Dispatch, a Skill for consulting external coding-agent CLIs and evaluating their responses within the current workflow.
The boss-skill creator explains why copying Skill files does not ensure hooks are connected. The project tracks installed files with a lockfile and checks expected files across target agents with an installation test matrix.
dontbesilent explains how to start with the dbskill toolkit: its onboarding conversation introduces the available Skills and helps users choose an entry for their current task.
stop-describing-ui trims explanatory text from generated interfaces and replaces vague labels with specific actions. Its rules retain critical consequences such as deletion, charges and customer notifications.
grim:lock-in helps resume work after a context switch by reconstructing the current project state, open loops, and three small next steps.
The PDB archive Skill retrieves experimental molecular structures, RCSB metadata and original mmCIF files with provenance. The creator’s science-data series shows a 3D demo of ten familiar proteins.