Skip to content

Releases: sohilpandya/skills

v1.2.1

18 Feb 17:42
@sohilpandya sohilpandya
v1.2.1
7d03571

Choose a tag to compare

View all tags
v1.2.1 Latest
Latest

What's new

  • Add demo video showcasing snap-context in action (intro scene, screenshot flash, conversion, tagline)
  • Embed demo video at top of README
  • Include Remotion source in demo/ for regenerating the video
Assets 2
Loading

v1.2.0

18 Feb 16:48
@sohilpandya sohilpandya
v1.2.0
369be68

Choose a tag to compare

View all tags
v1.2.0

What's Changed

  • Renamed repo from snap-context to skills to support multiple skills
  • Rewrote top-level README as a multi-skill index with installation options
  • Added skill-level README at skills/snap-context/README.md with full docs
  • Updated install paths to use --skill flag for targeted installation
Loading

v1.1.0

17 Feb 18:26
@sohilpandya sohilpandya
v1.1.0
9be98bc

Choose a tag to compare

View all tags
v1.1.0

What's changed

  • Broadened skill trigger description to cover more natural phrasings
    • Added leading verb cluster: "Analyze, describe, read, or extract"
    • Added noun synonyms: "pic", "screen grab"
    • Grouped triggers by intent: questions, review requests, extraction requests, attachment descriptions
    • Explicitly covers phrases like "what can you see", "review these", "I pasted this"
    • Works with single or multiple images
Loading

v1.0.0 — Initial Release

17 Feb 17:57
@sohilpandya sohilpandya
v1.0.0
e9f2284

Choose a tag to compare

View all tags
v1.0.0 — Initial Release

First official release of snap-context — a Claude Code skill that converts screenshots into clean, structured markdown.

Features

Screenshot to Structured Markdown

Analyze screenshots of UI data and convert them into well-formatted markdown. Supports 7 structure types:

  • Table — pipe-delimited markdown tables
  • Form — bullet lists with bolded labels
  • Card — headed sections for side-by-side content blocks
  • Code — fenced code blocks with language detection
  • Dialog — blockquoted modals with title and buttons
  • Hierarchy — nested lists preserving indent and bullet types
  • Plain Text — paragraph-formatted text

Context-Preserving Architecture

All image analysis is delegated to subagents via the Task tool, keeping image tokens out of the main conversation context.

Parallel Multi-Image Support

When multiple screenshots are shared, each is processed by a separate subagent in parallel for faster results.

Broad Trigger Coverage

Skill triggers on a wide range of phrases: "check this screenshot", "see what I've attached", "what's in this image", "extract text from this", and more — not just explicit extraction requests.

Installing

npx skills add sohilpandya/snap-context

Updating

# If installed via git clone or submodule
git pull

# If installed via npx skills
npx skills add sohilpandya/snap-context
Loading