What it is
This repo describes Social Post Helper as a lightweight Next.js app that helps people upload media, read available metadata, add a little context, and turn that into drafts they can edit, copy, or export.
Web app / MVP / local-first
Turn the clues already inside your media into a usable first draft.
Social Post Helper reads dates, dimensions, device fields, GPS when available, and your own notes, then builds editable starting points for Instagram, Facebook, and YouTube without leaning on AI.
- Status
- MVP
- Format
- Static-friendly web app
- Method
- Metadata + templates + edits
Why it exists
A calmer first pass for social posting
Why it matters
The useful part is not pretending to understand the image. It is taking repetitive setup work off the table so the user starts from a structured draft instead of a blank box.
What it reduces
- Blank-page caption writing
- Manual retyping of dates, locations, and media context
- Inconsistent copy across Instagram, Facebook, and YouTube
- Tooling creep for people who do not want AI in the loop yet
Core surface
Built like a contact sheet, not a black box
Metadata-first intake
Reads created date, GPS when present, camera/device fields, file type, resolution, and video duration.
Rule-based drafting
Generates platform-specific drafts through deterministic templates instead of model output.
Manual context where it counts
Add category, what happened, audience, and notes so the drafts stay useful without pretending to see the media.
Platform coverage
Includes Instagram caption, Facebook post, YouTube title options, description, upload summary, hashtags, alt text, and a short caption.
Local memory
Preferences and the last session snapshot are stored in the browser so the workflow can reopen without accounts.
Careful location handling
Reverse geocoding is optional, with a coordinate fallback when location lookup is unavailable or disabled.
Field sequence
How the repo says the product should be used
-
1. Bring in the media
Upload one or more photos and, for this MVP, up to one video.
-
2. Read the detected signal
Review dates, dimensions, GPS availability, media mix, and the structured summary before writing.
-
3. Add intent
Pick a category, add what happened, note who the post is for, and choose tone and platform preferences.
-
4. Edit the first draft
Adjust the generated text, copy what works, and export JSON or plain text if needed.
Travel or outing
GPS and date can become a clean location-aware starting point instead of another post written from scratch.
Project progress
Multiple photos and a manual note become a short progress update without needing a new writing sprint.
YouTube upload recap
The MVP stays metadata-driven for video and turns filename, notes, date, and location into a basic publishing pack.
Visual reference
No repo screenshots were available, so this export uses calm product diagrams instead
Proof and limits
Specific claims only
No AI in this MVP
The README explicitly says no LLMs, no cloud AI, no OCR, and no paid AI APIs.
Static export path documented
The repo includes export-oriented Next.js settings and a GitHub Pages workflow for static deployment.
Validation recorded on 2026-03-17
Local checks passed for tests, TypeScript, and lint at the time this export package was built.
Known limits stay visible
No transcript extraction, no OCR, no direct posting, no accounts, and session restore does not reattach original file blobs.
Repo history also shows focused compatibility work for HEIC/HEIF handling and Safari/macOS Photos picker behavior. This matters because upload reliability is part of the product, not an afterthought.
FAQ
Plain answers
Is Social Post Helper using AI?
No. The documented MVP is deterministic and metadata-first, with future AI integration isolated for later work.
What does it actually read from the media?
When available, it reads created date, GPS, camera or device fields, file type, image or video dimensions, and video duration.
Where does the draft data live?
Preferences and the last session snapshot are stored in browser localStorage in the MVP.
Can it restore the original uploaded files later?
No. The documented session restore brings back metadata context and drafts, not the original file blobs.
What kinds of media are in scope?
Photos are the main path, with one video supported for metadata-driven draft generation in this MVP. HEIC preview depends on browser support.
Does it post directly to social platforms?
No. Direct publishing, OAuth flows, accounts, scheduling, transcript extraction, and OCR are all out of scope for the current MVP.
Final note
Useful before it gets ambitious
Social Post Helper earns its place by making the first draft smaller, cleaner, and more honest. It is a good fit for people who want structure from their own media before they want AI to rewrite everything.