| name | social-media-publishing |
| description | Multi-channel social publishing workflow plus blog-post publication prep and channel CLIs for Reddit, LinkedIn, X, YouTube, Instagram, and TikTok. Use when Codex needs to publish or distribute a blog post, video, launch, or visual explainer; prepare blog assets; decide between gallery, image, self, link, or native video format; publish the source post before distribution; post to Reddit; publish LinkedIn text/link/image/video posts; post X text/video posts; or prepare Instagram/TikTok posts through dry-run channel CLIs. |
Social Media Publishing
Overview
Use this skill to package publishing and distribution work as a reusable workflow instead of repo-specific one-offs.
Keep campaign assets in the active project. Keep durable repo-specific publishing conventions in the owning repo. The current bundled channel helpers cover Reddit, LinkedIn, X, Modal-backed YouTube uploads, and initial dry-run scaffolds for Instagram and TikTok.
Workflow
- Decide whether the content needs to be published at the source first, for example in a blog repo.
- Keep campaign state in the active repo, not in this skill.
- Choose the post shape with the user first: gallery, self, link, or image.
- Inspect subreddit flairs before posting when flair is required or likely to matter.
- Prefer a plan file for repeatable or multi-step posts.
- Dry-run the plan before live submission when paths, long comments, or strict communities are involved.
- After posting, update the project-local tracker with the post URL, comment URL, and moderation outcome.
- For source-post workflows, default to publishing the post publicly (set
hidden: false in frontmatter unless the user explicitly asks for a private/draft post).
Blog-backed publishing
If the user asks to publish the source post itself before distribution, read:
references/blog/publishing.md
Keep durable repo-specific publishing mechanics in the owning repo and in the reference file, not in this top-level skill body.
Reddit
Use the bundled CLI at scripts/reddit/cli.py.
For live commands and plan structure, read:
references/reddit/workflow.mdreferences/reddit/plan-schema.md
If the content is Codex / App Server / coding-agent-internals related and you expect repeated Reddit distribution, also read:
references/reddit/topics/codex.md
Core commands:
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/reddit/cli.py status
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/reddit/cli.py list-flairs --subreddit OpenAI
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/reddit/cli.py list-submissions --max-items 20 --days 7
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/reddit/cli.py comment --post-url https://reddit.com/r/OpenAI/comments/abc123/example/ --text-file /abs/path/comment.md --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/reddit/cli.py submit-plan --plan /abs/path/post-plan.json --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/reddit/cli.py submit-plan --plan /abs/path/post-plan.json
The CLI supports:
- flair discovery
- gallery submission
- image submission
- link submission
- self-post submission
- optional first-comment posting
- standalone commenting on existing posts
- authenticated submission history lookup
LinkedIn
Use the bundled LinkedIn CLI at scripts/linkedin/cli.py for local LinkedIn publishing.
Read first:
references/linkedin/posting.mdreferences/linkedin/copy.md
If LinkedIn app product access changes, company pages are involved, or analytics/read-back is needed:
references/linkedin/community-management.md
Only if setup or re-auth is needed:
references/linkedin/auth.md
Current supported flow:
- local OAuth authorization
- Community Management API scope presets and capability probes
- runtime/auth inspection via
status
- identity check
- text posts
- article or URL shares
- single-image posts
- single-video posts
- multi-image posts
- comments on posts
- comment read-back
- organization/page role lookup
- member post and video analytics
- machine-readable JSON output by default, plus optional
--plain inspection mode
- stderr-only progress for long-running video uploads via
--progress
Core commands:
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py status
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py authorize
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py authorize --scope-preset community
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py community-status
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py whoami
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py post --text-file /abs/path/body.txt --url https://example.com/post --title "Post title" --description "Short description" --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py post-image --text-file /abs/path/body.txt --image /abs/path/cover.jpg --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py post-video --text-file /abs/path/body.txt --video /abs/path/video.mp4 --title "Public video title" --thumbnail /abs/path/thumbnail.png --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py post-video --text-file /abs/path/body.txt --video-url https://example.com/video.mp4 --title "Public video title" --thumbnail /abs/path/thumbnail.png --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py post-images --text-file /abs/path/body.txt --image /abs/path/slide-1.jpg --image /abs/path/slide-2.jpg --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py comment --post-urn urn:li:ugcPost:... --text-file /abs/path/comment.txt --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py list-comments --post-urn urn:li:ugcPost:...
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py organization-acls
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py member-post-analytics --post-urn urn:li:share:... --metric IMPRESSION
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py member-video-analytics --post-urn urn:li:ugcPost:... --metric VIDEO_PLAY
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/linkedin/cli.py --json list-posts --count 5
For polished LinkedIn native video posts, always pass --title; LinkedIn shows it as the media title below the player. Pass --thumbnail when a custom cover is available. If --title is omitted, the helper leaves the media title unset instead of deriving one from a temporary download filename.
This helper uses machine-local generated secrets under ~/.secrets/linkedin/ and should stay one-user until the workflow is more mature.
Instagram
Use the bundled Instagram CLI at scripts/instagram/cli.py for local Instagram publishing preparation.
Read first:
references/instagram/posting.md
Current supported flow:
- runtime/config inspection via
status
- dry-run validation for image posts, video/Reel posts, and carousel posts
- machine-readable JSON output by default, plus optional
--plain inspection mode
- live publishing intentionally disabled pending Meta app/account/token setup
Core commands:
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/instagram/cli.py status
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/instagram/cli.py post-image --text-file /abs/path/caption.txt --image-url https://example.com/image.jpg --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/instagram/cli.py post-video --text-file /abs/path/caption.txt --video-url https://example.com/video.mp4 --reel --share-to-feed --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/instagram/cli.py post-carousel --text-file /abs/path/caption.txt --media-url https://example.com/1.jpg --media-url https://example.com/2.jpg --dry-run
TikTok
Use the bundled TikTok CLI at scripts/tiktok/cli.py for local TikTok publishing preparation.
Read first:
references/tiktok/posting.md
Current supported flow:
- runtime/config inspection via
status
- dry-run validation for video posts and photo-mode posts
- machine-readable JSON output by default, plus optional
--plain inspection mode
- live publishing intentionally disabled pending TikTok developer app/OAuth/API review setup
Core commands:
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/tiktok/cli.py status
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/tiktok/cli.py post-video --text-file /abs/path/caption.txt --video-url https://example.com/video.mp4 --privacy SELF_ONLY --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/tiktok/cli.py post-video --text "Short caption" --source FILE_UPLOAD --video /abs/path/video.mp4 --ai-generated --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/tiktok/cli.py post-photos --title "Photo title" --description-file /abs/path/description.txt --photo-url https://example.com/1.jpg --photo-url https://example.com/2.jpg --cover-index 0 --privacy SELF_ONLY --dry-run
YouTube
Use the bundled YouTube CLI at scripts/youtube/cli.py for local YouTube publishing through the existing Modal upload engine.
Read first:
references/youtube/posting.md
Current supported flow:
- runtime/config inspection via
status
- upload from a public direct
--video-url through Modal
- upload from a local
--video by staging into the Modal cache volume, then calling Modal
- intentionally small upload surface: video, title, description, optional thumbnail, optional privacy
- machine-readable JSON output by default, plus optional
--plain inspection mode
- stderr-only progress via
--progress
Core commands:
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/youtube/cli.py status
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/youtube/cli.py upload-video --video /abs/path/video.mp4 --title "Video title" --description-file /abs/path/description.md --privacy private --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/youtube/cli.py --progress plain upload-video --video /abs/path/video.mp4 --title "Video title" --description-file /abs/path/description.md --privacy unlisted
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/youtube/cli.py upload-video --video-url https://example.com/video.mp4 --title "Video title" --description-file /abs/path/description.md --privacy unlisted --dry-run
YouTube is Modal-backed by design. The CLI is personal-account-first: credentials id ADITHYAN, notify subscribers on, made-for-kids off, embeddable on. Do not import WIN or duplicate the local YouTube uploader in this skill unless Modal becomes a proven bottleneck.
X
Use the bundled X CLI at scripts/x/cli.py for local X posting.
Read first:
references/x/posting.mdreferences/x/copy.md
Only if setup is still missing:
Current supported flow:
- runtime/auth inspection via
status
- text posts
- native video posts from a local file or public direct video URL
- machine-readable JSON output by default, plus optional
--plain inspection mode
- stderr-only progress for long-running video uploads via
--progress
Core commands:
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/x/cli.py status
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/x/cli.py post --text-file /abs/path/body.txt --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/x/cli.py post-video --text-file /abs/path/body.txt --video /abs/path/video.mp4 --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/x/cli.py post-video --text-file /abs/path/body.txt --video-url https://example.com/video.mp4 --dry-run
python3 ~/GitHub/agents/skills-source/owned/social-media-publishing/scripts/x/cli.py post --text-file /abs/path/body.txt
Packaging Rules
- Keep the skill reusable and the campaign local.
- Draft assets go in
<repo>/tmp/social/<channel>/<slug>/ (gitignored, throwaway). Do not put drafts in capture/ or other tracked folders.
- Keep credentials in environment variables or the user's authenticated tools, not in project files.
- Use absolute paths in ad hoc commands when possible.
- Use relative paths inside plan files when the plan lives next to the assets.
- Prefer plan files over long shell commands once the post has more than a title plus one asset.
- If the social post depends on a live blog URL, publish and verify the blog post first.
Resources
scripts/reddit/cli.py: Reddit CLI entrypoint.scripts/reddit/: self-contained Reddit helpers and models.references/blog/publishing.md: blog publication workflow before distribution.references/reddit/workflow.md: operational Reddit workflow.references/reddit/plan-schema.md: portable Reddit plan-file contract.scripts/youtube/cli.py: Modal-backed YouTube upload CLI.references/youtube/posting.md: YouTube upload workflow and route contract.scripts/linkedin/cli.py: local LinkedIn posting CLI.references/linkedin/posting.md: LinkedIn posting setup and usage.references/linkedin/copy.md: LinkedIn copy defaults and reusable post baselines.references/linkedin/community-management.md: Community Management API access, tiers, and next upgrade path.scripts/x/cli.py: local X posting CLI.references/x/posting.md: X posting setup and usage.references/x/copy.md: X copy defaults.scripts/instagram/cli.py: Instagram dry-run publishing prep CLI.references/instagram/posting.md: Instagram setup, provider constraints, and dry-run usage.scripts/tiktok/cli.py: TikTok dry-run publishing prep CLI.references/tiktok/posting.md: TikTok setup, provider constraints, and dry-run usage.
