菜单
Cross-post LinkedIn content to X (Twitter). Use when user says "post to x", "cross-post to x", "tweet this", "share on x", "post on twitter", or wants to adapt a LinkedIn post for X/Twitter. Handles text shortening, media upload (images and videos), and posting via the X API v2 + v1.1 media endpoint.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Meta-skill that analyzes past session transcripts to derive new skills, CLAUDE.md rules, and MEMORY.md entries from real usage patterns. Use when Use when user says "session learn", "derive skills", "learn from sessions", "what patterns", "analyze sessions", "extract rules", "what should I add to claude md", "improve my setup", or "what have I been doing repeatedly".
Scan running Claude sessions to see what other agents are working on. Use when asked "what are the other agents doing", "check other sessions", "what's running", "scan agents", "who's working on what", or before picking up new work to avoid overlap.
Find restaurants and food delivery options near a location. Searches Swiggy for restaurants, cross-references Google Maps ratings, and recommends the best options. Learns food preferences over time.
Geographic data validator for aviation and travel applications. Use when user says "geo audit", "check airports", "validate routes", "geo skill", "check geographic data", "verify iata codes", "check distances", or when reviewing any code/data that involves airport codes (IATA/ICAO), city-to-airport mappings, country codes (ISO 3166), flight distances, flight time estimates, flag URLs, or route feasibility. Also use for timezone validation, coordinate checks, and geopolitical accuracy (country names, disputed territories, naming conventions).
A set of resources to help write all kinds of internal communications, using the formats that your company likes to use. Claude should use this skill whenever asked to write some sort of internal communications (status reports, leadership updates, 3P updates, company newsletters, FAQs, incident reports, project updates, etc.).
Negotiation analyst for high-stakes deals: co-founder exits, business separations, contract terms, salary negotiations, investor terms. Use when the user asks to: (1) analyze a counterparty's position or message, (2) score/review a draft message before sending, (3) draft a counter-offer, (4) structure BATNA/ZOPA analysis, (5) track concessions across rounds, (6) detect traps or information leakage in proposed terms, (7) prepare for a negotiation call or meeting. Triggers: "score this message", "analyze their position", "what's my BATNA", "review before I send", "draft a counter", "is this a trap", "negotiate", "negotiation", "counter-offer", "deal terms".
| name | post-to-x |
| description | Cross-post LinkedIn content to X (Twitter). Use when user says "post to x", "cross-post to x", "tweet this", "share on x", "post on twitter", or wants to adapt a LinkedIn post for X/Twitter. Handles text shortening, media upload (images and videos), and posting via the X API v2 + v1.1 media endpoint. |
Cross-post LinkedIn content to X, adapting text and attaching media.
X API credentials must be set in ~/.claude/skills/post-to-x/.env:
X_API_KEY=your_api_key
X_API_SECRET=your_api_secret
X_ACCESS_TOKEN=your_access_token
X_ACCESS_TOKEN_SECRET=your_access_token_secret
These come from the X Developer Portal (developer.x.com). You need a project with "Read and Write" permissions. The access token must be generated with read+write scope.
Look for the post text in the current directory or the directory the user references:
post-copy-final.txt or post-copy*.txt for text*.mp4, *.mov, *.webm for video*.jpg, *.jpeg, *.png for imagesthumbnail*.jpg, thumbnail*.png for thumbnail imagesIf the user provides text directly, use that instead.
LinkedIn posts are long-form. X has character limits:
Strategy:
If the post is too rich for a single tweet, propose a thread (2-4 tweets max).
Always show the adapted text to the user for approval before posting.
python3 ~/.claude/skills/post-to-x/scripts/post_to_x.py upload-media <file_path>
This returns a media_id to use in Step 4.
For video: the script handles chunked upload (INIT, APPEND, FINALIZE, STATUS polling). For images: simple upload.
Size limits:
# Text only
python3 ~/.claude/skills/post-to-x/scripts/post_to_x.py post --text "Your tweet text"
# With media
python3 ~/.claude/skills/post-to-x/scripts/post_to_x.py post --text "Your tweet text" --media-ids "media_id_1,media_id_2"
# Thread
python3 ~/.claude/skills/post-to-x/scripts/post_to_x.py thread --texts "Tweet 1|||Tweet 2|||Tweet 3"
# Thread with media on first tweet
python3 ~/.claude/skills/post-to-x/scripts/post_to_x.py thread --texts "Tweet 1|||Tweet 2" --media-ids "media_id_1"
# Dry run (show what would be posted without actually posting)
python3 ~/.claude/skills/post-to-x/scripts/post_to_x.py post --text "Your tweet" --dry-run
The script prints the tweet URL on success. Confirm the post looks correct.
LinkedIn (long):
I genuinely stopped using PowerPoint.
Spending days aligning boxes feels insane when you can build a whole app in the same time.
So I made an app for building slides.
Three prompts. Full deck. 2 minutes.
No Figma. No templates. Just prompts.
Open source. Free forever.
DM "DECK" if you want early access.
X (short):
I stopped using PowerPoint and built an app that generates full slide decks from prompts.
3 prompts. Full deck. 2 minutes.
Open source, free forever.
github.com/yourorg/yourrepo
.env--dry-run flag.