Reel2MD 0.1.4

Reel2MD 0.1.4

Reel2MD 0.1.4 is a Community Directory build verification maintenance release.

Highlights

• Made the repository root build self contained for clean Community Directory build verification.
• Removed the build time dependency on a nested plugin/node_modules installation.
• Kept the generated Obsidian release assets unchanged in structure: main.js, manifest.json, and styles.css.
• Clarified README disclosures for Node.js filesystem access and local child process execution.
• No functional changes were made to Reel processing, transcription, queue behavior, or network behavior.

Validation

Before release, 0.1.4 is intended to pass:

• Python regression tests.
• A clean root npm ci followed by npm run build.
• Root/plugin manifest synchronization.
• Clean wheel installation and CLI version verification.
• Local Obsidian smoke testing.
• Obsidian Community readiness checks.

The GitHub native release workflow rebuilds the public artifacts from main, verifies checksums, creates provenance attestations for the Obsidian release assets, and only then creates the GitHub release.

Reel2MD 0.1.3

Reel2MD 0.1.3

Reel2MD 0.1.3 is a maintenance and Community submission hardening release built on top of 0.1.2.

Highlights

• Added the Obsidian version compatibility mapping required for clean Community release metadata.
• Strengthened release provenance with GitHub artifact attestations for the Obsidian release assets.
• Aligned root development dependencies so Community scanner/build tooling can resolve the Obsidian and TypeScript types consistently.
• Normalized child process error values before rejection for safer error handling.
• Added the A RichDir project identity link to the README and Reel2MD settings.
• Kept the plugin desktop only, local first, and without analytics or telemetry.

Validation

The release candidate is prepared to pass:

• Python regression tests.
• TypeScript production build.
• Root plugin build.
• Root/plugin manifest synchronization.
• Clean wheel installation and CLI version verification.
• Local Obsidian smoke testing before the branch is pushed.
• Obsidian Community readiness checks after the release prep PR is opened.

The GitHub-native release workflow rebuilds public artifacts from main, verifies checksums, creates provenance attestations for main.js, manifest.json, and styles.css, and only then creates the GitHub release.

Reel2MD 0.1.2

Reel2MD 0.1.2

Reel2MD 0.1.2 focuses on Obsidian Community Plugins readiness, settings modernization, release validation, and multilingual transcription support.

Highlights

• Added official Obsidian Community readiness validation.
• Migrated the settings UI to Obsidian's declarative settings API.
• Resolved all current Community scanner findings.
• Added English only and multilingual transcription model selection.
• Multilingual mode uses small and automatically detects the spoken language.
• English only mode uses small.en.
• Added safer child process cleanup when the plugin unloads.
• Added normalized Obsidian vault-path handling.
• Added production minification and a reproducible root plugin build path.
• Improved README and installation documentation for 0.1.2.

Validation

Before release, 0.1.2 passed:

• Python regression tests.
• TypeScript production build.
• Root plugin build.
• Clean wheel installation.
• Local dependency checks.
• Instagram/Hugging Face network diagnostics.
• Obsidian settings UI and conditional-field tests.
• English transcription end to end testing.
• French multilingual auto detection and transcription end to end testing.

Reel2MD 0.1.1

Reel2MD 0.1.1

This is the first public release of Reel2MD intended for normal use. This release focuses on settings UX, local diagnostics, release safety, and Obsidian Community Plugins readiness.

Highlights

• Reorganized and simplified Obsidian settings.
• Conditional settings for transcription, authenticated fallback, proxy configuration, and media retention.
• Added local dependency status for Reel2MD, ffmpeg, yt-dlp, faster-whisper, and huggingface-hub.
• Added optional live job output inside Obsidian.
• Enforced a 2-second minimum request spacing with a configurable 2–60 second maximum.
• Added CLI safety validation for retained media.
• Added reel2md deps for local dependency diagnostics.
• Added support for queue-note paths with or without .md.
• Improved documentation and external-access disclosure.
• Updated the Obsidian plugin ID to reel-to-md.
• Added a repository-root manifest.json for Community Plugins submission readiness.

Validation

Release artifacts were built from commit:

ecc0a965bb2ad5f5da3ad67b4690d8c8312ca585

The release passed:

• Python regression tests
• TypeScript production build
• clean wheel installation
• dependency checks
• Instagram/Hugging Face network diagnostics
• Obsidian plugin smoke test
• end-to-end queue processing
• SHA-256 artifact verification

Reel2MD 0.1.0

Reel2MD 0.1.0

First public alpha release of Reel2MD.

Reel2MD converts supported Instagram video posts into structured Markdown notes, with optional local speech transcription and an Obsidian desktop plugin for configuration and queue processing.

Highlights

• Supports Instagram /reel/, /reels/, /p/, and legacy /tv/ URL forms.
• Canonicalizes tracking variants and deduplicates by Instagram shortcode.
• Extracts captions and source metadata with yt-dlp.
• Uses stable source IDs for filenames and writes the human-readable title to Obsidian aliases.
• Optional local transcription with faster-whisper.
• Streams media to ffmpeg; video does not need to be kept permanently.
• Optional video and transcription-audio retention.
• Optional authenticated browser-session fallback.
• Configurable request spacing and rate-limit stop behavior.
• Proxy modes: inherit system, no proxy, or manual HTTP/HTTPS proxy.
• Optional HF_HUB_DISABLE_XET=1 support for Hugging Face compatibility.
• Test setup and Test network checks in the Obsidian plugin.
• Clear handling for photo-only Instagram /p/ posts.
• Regression coverage for URL parsing, canonicalization, deduplication, aliases, and non-video post handling.

Requirements

• Desktop Obsidian for the plugin.
• Python 3.10+.
• ffmpeg.
• Network access to Instagram.
• Network access to Hugging Face when the selected transcription model is not cached.

Release assets

Obsidian plugin:

• main.js
• manifest.json
• styles.css
• reel2md-0.1.0-plugin.zip

Python CLI:

• reel2md-0.1.0-py3-none-any.whl
• reel2md-0.1.0.tar.gz

Integrity:

• SHA256SUMS.txt

Alpha status

This is an early public alpha extracted from a working personal workflow. Instagram behavior and endpoints can change, and broader OS/environment testing is still limited.

Use Reel2MD only for content you are permitted to access and process.