✨ Features Overview

Indexly is a modular, offline-first indexing and search system for structured and unstructured data.

Indexly is a local-first file intelligence engine.
It indexes, organizes, analyzes, and inspects data — entirely offline, fully auditable, and designed for power users.

This page describes what Indexly can do.
For step-by-step usage, see the Usage Guide.


🔍 Search Engine (Core Capability)

Indexly provides multiple complementary search engines over your local files:

  • Full-text search (SQLite FTS5) over content and metadata
  • Boolean logic: AND, OR, NOT, NEAR
  • Phrase and proximity search
  • Regex search for structural and pattern-based queries
  • Fuzzy matching for imperfect input
  • Ranking and scoring based on relevance

Search operates over indexed content, not raw files — enabling instant queries across large datasets.

→ See: Search Documentation


🗂 Semantic Indexing

Indexly builds a semantic index, not just a filename cache:

  • Extracts text from PDFs, Office files, emails, code, and logs
  • Indexes metadata (author, title, dates, EXIF, document info)
  • Applies ignore rules via .indexlyignore
  • Supports incremental updates and cache refresh

Indexing is the foundation for fast, repeatable analysis.

→ See: Indexing Files


🏷 Tag Intelligence

Indexly supports two tagging layers:

  • Manual tags (user-defined, persistent)
  • Virtual tags (regex-based, derived from content)

Tags integrate directly into search and filtering without modifying files.

→ See: Tagging


📋 Lister & Auditing

Lister is Indexly’s read-only inspection layer:

  • Reads organizer logs or synthesized logs
  • Filters by category, extension, date, duplicates
  • No filesystem mutation
  • Ideal for audits, compliance, and verification

→ See: Lister


🔁 Organizer & Observability

Organizer classifies files into structured layouts and emits machine-readable JSON logs.

These logs power:

  • Lister
  • Audits
  • Re-runs
  • External automation

Organizer is optional — but when used, it turns Indexly into a fully observable system.

→ See: Organizer


📊 Data Analysis

Indexly includes built-in analyzers for structured data:

  • CSV, JSON/NDJSON, XLSX, XML
  • SQLite and Parquet inspection
  • Statistical summaries and schema insights

→ See: Data Analysis

🧮 Statistical Inference (CSV Only)

Indexly can perform rigorous statistical inference directly on CSV datasets:

Core Capabilities

  • 📊 Correlation Analysis – Pearson (Fisher Z CI), Spearman, lag correlation, full correlation matrices
  • 🧪 Parametric Tests – Independent and paired t-tests, one-way ANOVA, Tukey post-hoc
  • 📈 Regression Models – OLS with interaction terms and mixed-effects modeling
  • 🛡 Assumption-Aware – Automatic rerouting to nonparametric tests when assumptions fail
  • 🔁 Bootstrap Support – Optional bootstrap confidence intervals and coefficient estimation
  • 📄 Structured Export – Export formatted inference reports to Markdown or PDF

→ See: Inference Documentation


📦 Comparison & Diffing

Indexly can compare:

  • Files
  • Folders
  • Structured data

With similarity scoring, context folding, and scriptable exit codes.

→ See: Compare


🛡 Architecture Principles

  • Local-only execution
  • No background services
  • No network calls
  • Fully inspectable outputs
  • Deterministic behavior

Indexly is designed for trust, traceability, and control.


🔗 Next Steps