Crate selah 

Selah — AI-native screenshot capture, annotation, and redaction library for AGNOS.

Re-exports

pub use ai::OcrResult;

pub use ai::RedactionSuggestion;

pub use ai::SmartCropSuggestion;

pub use ai::extract_text_regions;

pub use ai::suggest_redactions;

pub use ai::suggest_smart_crop;

pub use annotate::AnnotationCanvas;

pub use annotate::AnnotationLayer;

pub use annotate::convert_format;

pub use capture::CaptureClient;

pub use capture::CaptureResponse;

pub use core::Annotation;

pub use core::AnnotationKind;

pub use core::CaptureRegion;

pub use core::CaptureSource;

pub use core::Color;

pub use core::ImageFormat;

pub use core::Monitor;

pub use core::RedactionTarget;

pub use core::Screenshot;

pub use core::derive_output_path;

pub use core::xml_escape;

pub use daimon::DaimonClient;

pub use daimon::DaimonConfig;

pub use daimon::HooshClient;

pub use daimon::HooshConfig;

pub use error::SelahError;

pub use geometry::Rect;

pub use history::HistoryEntry;

pub use history::HistoryStore;

Modules

ai

AI features for Selah: OCR, PII detection, smart crop suggestions.

annotate

Annotation engine for Selah.

capture

Screen capture API client for Selah — connects to daimon.

core

Core types and primitives for the Selah screenshot tool.

daimon

Daimon and Hoosh integration for Selah.

error

Error types for Selah.

geometry

Geometry types backed by hisab.

history

Screenshot history — persisted as JSON-lines.

Structs

Vec2

Re-export hisab’s 2D vector type for use as points and coordinates. A 2-dimensional vector.

Functions

annotate_image

Apply annotations to an image and return the encoded bytes.

decode_image_data

Decode base64-encoded image data from a capture response.

redact_image

Detect PII in an image and produce redacted bytes. Returns (redacted image bytes, list of suggestions).