Updates

Changelog

All releases, features, and fixes.

Latest

v0.6.0

Releases

Changes

149

Release History

22 versions

0.6.0CurrentFeb 18, 2026

HTTP API Server & Local Import Resolution

New HTTP API server alternative to MCP stdio, local import path resolution, Python stdlib detection, and symbol parent tracking.

FeatureHTTP API server — Alternative to MCP stdio with `--http-port`, endpoints: `/health`, `/stats`, `/search`, `/find-refs`, `/deps`, `/explain`

FeatureHTTP middleware — 1MB body limit, 50 concurrent requests, CORS configurable via `--cors-origin`

FeatureLocal import resolution — Resolution of local import paths to actual files on disk (JS/TS, Python, Rust, Go)

Feature`resolved_path` column — Dependencies now store the resolved path, improving `find_refs` accuracy

FeaturePython stdlib detection — Accurate classification of Python standard vs external imports (200+ modules, binary search)

FeatureSymbol parent tracking — Symbols now include their parent (e.g., method → struct/class)

FeatureDockerfile — Multi-stage Docker image for deployment (Railway-ready)

ImprovementSchema migration v3→v4 — Automatic incremental migration (adds `resolved_path` column, no reindex required)

ImprovementAuto-indexer and CLI `index` command use local import resolution

FixCorrect git clone URL in Dockerfile

FixResolve clippy `module_inception` warning in HTTP tests

FixBump Rust version in Dockerfile to support edition 2024 and let-chains

0.5.2Feb 10, 2026

Security Hardening

Major security release with fixes for ReDoS, path traversal, and input validation vulnerabilities across the codebase.

Fix[HIGH] Fixed ReDoS vulnerability - user input is now escaped with `regex::escape()` in `TextSearcher::search()` before regex compilation

Fix[HIGH] Updated `bytes` crate 1.11.0 → 1.11.1 to fix integer overflow in `BytesMut::reserve` (RUSTSEC-2026-0007)

Fix[HIGH] Text search walker now uses `hidden(true)` and `should_exclude_entry` filtering, preventing reads from `.env`, `.git/`, and other sensitive directories

Fix[MEDIUM] Fixed path traversal in `read_file_lines()` - paths are now canonicalized and verified to stay within the project root

Fix[MEDIUM] Added input validation (empty, length ≤ 500, limit ≤ 1000) to `semantiq_find_refs`, `semantiq_explain`, and `semantiq_deps` MCP handlers

Fix[MEDIUM] Added path traversal rejection (`..`) in `semantiq_deps` file path parameter

Fix[MEDIUM] `resolve_project_root()` now canonicalizes paths to normalize `..` components and symlinks

Fix[LOW] FTS5 query escaping now strips null bytes and control characters

Fix[LOW] Query expansion limited to 10 terms to prevent amplification attacks

Fix[LOW] MCP error messages sanitized to avoid leaking internal file paths

Fix[LOW] Version check HTTP response limited to 10KB to prevent memory exhaustion

Fix[LOW] Poisoned mutex recovery in `DistanceCollector` now logs warnings instead of silently continuing

0.5.0Jan 31, 2026

Adaptive ML Thresholds

Automatic calibration of semantic search thresholds per programming language, new CLI commands, and codebase refactoring.

FeatureAdaptive ML thresholds - automatic calibration of semantic search thresholds per programming language

FeatureBootstrap mode - collects 100% of distance observations until 500 samples

FeatureProduction mode - switches to 10% sampling after bootstrap completes

FeatureAuto-calibration - triggers automatically when bootstrap completes

FeaturePercentile-based thresholds - uses p90 for max_distance, p10 for min_similarity

FeaturePer-language calibration with fallback cascade (language → global → defaults)

FeatureNew `calibrate` CLI command - manual threshold calibration with `--dry-run` option

FeatureML stats in `stats` command - shows bootstrap progress, observations per language, calibrated thresholds

FeatureNew database tables - `distance_observations` and `threshold_calibration` for ML data

FeatureCI workflows for `dev` branch - tests, Clippy, format checks, and multi-platform builds

ImprovementRefactored `store.rs` (2108 lines → 8 modules) - better code organization with dedicated modules for files, symbols, chunks, dependencies, observations, calibrations, and tests

ImprovementRefactored `engine.rs` (1049 lines → 5 modules) - cleaner architecture with dedicated modules for search, threshold, analysis, and tests

BreakingSchema version bumped to 3 - triggers automatic reindex on upgrade

0.4.0Jan 28, 2026

JSON Logging & MCP Tests

Structured JSON logging throughout the codebase, MCP test coverage, and CI/security workflows.

FeatureJSON logging support - structured logging throughout the codebase

FeatureJSON logging by default for `serve` command - better integration with log aggregators

FeatureMCP tests - test coverage for MCP server functionality

FeatureCI and security workflows - automated testing and security scanning

Improvement`init-cursor` command is now language-agnostic - works with any project type

ImprovementUpdated `deny.toml` to v2 schema

0.3.4Jan 20, 2026

macOS Intel Support & Optional ONNX

Restored macOS Intel (x86_64) support and made ONNX feature optional for broader platform compatibility.

FeaturemacOS Intel (x86_64-apple-darwin) support restored - binary now available for Intel Macs

FeatureCI build workflow - new build.yml for testing builds on push/PR without publishing

ImprovementONNX feature now optional - `--features onnx` required on supported platforms (Apple Silicon, Linux, Windows)

ImprovementIntel Mac builds use StubEmbeddingModel (no ONNX) due to missing prebuilt binaries

ImprovementUpdated CI to use macos-15 runner for Intel Mac cross-compilation

0.3.3Jan 19, 2026

Search Filtering Options

New search filtering capabilities for semantiq_search with min_score, file_type, and symbol_kind parameters.

Feature`min_score` parameter for `semantiq_search` - minimum relevance score threshold (0.0-1.0, default: 0.35)

Feature`file_type` parameter for `semantiq_search` - filter results by file extension

Feature`symbol_kind` parameter for `semantiq_search` - filter by symbol type (function, class, etc.)

FeatureCLI flags `--min-score`, `--file-type`, `--symbol-kind` for `semantiq search` command

FeatureSmart default exclusions - automatically excludes non-code files from search

Feature`SearchOptions` struct in `semantiq-retrieval` with builder pattern for flexible configuration

0.3.2Jan 19, 2026

Gitignore Support in init-cursor

The init-cursor command now automatically handles .gitignore entries for Semantiq database files.

Feature`.gitignore` support in `init-cursor` - automatically adds Semantiq database entries

FeatureCreates `.gitignore` if not present

FeatureUpdates existing `.gitignore` preserving original content

FeatureSkips if entries already present (no duplication)

Feature3 new tests for `.gitignore` handling in `init_cursor.rs`

0.3.1Jan 19, 2026

Cursor/VS Code Init Command

New init-cursor command for automatic Cursor/VS Code configuration setup, plus centralized CLI utilities.

FeatureNew `init-cursor` command for Cursor/VS Code configuration setup

FeatureCreates `.cursor/rules/project.mdc` (general project guidelines)

FeatureCreates `.cursor/rules/semantiq.mdc` (Semantiq MCP tools usage)

FeatureCreates `.cursor/mcp.json` (MCP server configuration)

FeatureCreates `.cursorignore` (indexing exclusions)

FeatureCreates `.vscode/` config (settings, tasks, launch, extensions)

FeaturePreserves existing files (skip instead of overwrite)

ImprovementCentralized `DEFAULT_DB_NAME` and path resolution utilities in `common.rs`

ImprovementRefactored all CLI commands to use shared utilities

ImprovementCLI description now generic ("for a project" instead of "for a Rust project")

Feature7 new unit tests for `common.rs` and `init_cursor.rs`

0.3.0Jan 19, 2026

Vector Search & Auto-Indexing

sqlite-vec integration for semantic vector search, automatic initial indexing, and 6 new languages.

Featuresqlite-vec integration for vector similarity search (384-dim MiniLM-L6-v2 embeddings)

FeatureAutomatic initial indexing when MCP server starts (no more manual `semantiq index` required)

Feature6 new languages: HTML, JSON, YAML, TOML, Bash, Elixir (total: 15 languages)

Featureripgrep integration for fast regex text search via `TextSearcher`

FeatureNew `search_similar_chunks()` method for semantic vector search

FeatureNew `InitialIndexResult` struct for tracking initial indexing progress

Fix"Imported by" always empty in `semantiq_deps` - rewrote `get_dependents()` to match JS/TS import paths

FixImport path resolution now handles basename matching with multiple extensions

ImprovementSchema version bumped to 2 (triggers automatic reindex)

ImprovementAdded `chunks_vec` virtual table for sqlite-vec embeddings

Improvement`start_auto_indexer()` now runs `initial_index()` before watching for changes

ImprovementImproved dependency matching with multiple LIKE patterns and post-filtering

0.2.9Jan 19, 2026

Arrow Function Indexing

Improved TypeScript/JavaScript function detection - arrow functions and function expressions are now correctly classified as functions.

FixArrow functions (const fn = () => {}) now correctly indexed as function instead of variable

FixFunction expressions (const fn = function() {}) now correctly indexed as function

ImprovementAdded is_function_variable() helper to detect functions assigned to variables

ImprovementAdded arrow_function and lexical_declaration to chunk boundaries for TypeScript/JavaScript

ImprovementBumped PARSER_VERSION to 3 (triggers automatic reindex)

0.2.8Jan 18, 2026

Security Hardening

Critical security improvements including checksum verification, path traversal protection, and memory exhaustion prevention.

FixCRITICAL: Added SHA-256 checksum verification for ONNX model downloads (TOFU + hardcoded support)

FixCRITICAL: Added path traversal protection with canonicalization in validate_path()

FixHIGH: Added MAX_AST_DEPTH=500 recursion limit in parser to prevent stack overflow attacks

FixHIGH: Added safe_slice() function to prevent panic on invalid byte indices

FixHIGH: Changed model directory fallback from "." to system temp dir (prevents writes to unexpected locations)

FixHIGH: Added pagination for get_chunks_with_embeddings() to prevent memory exhaustion DoS

FixHIGH: Reduced download size limit from 500MB to 100MB

FixHIGH: Added restrictive file permissions (0600 on Unix) for downloaded models and database

FixMEDIUM: Added explicit symlink handling (follow_links(false)) to prevent escape from project root

ImprovementRefactored download_file() with connection timeouts (30s connect, 5min global)

ImprovementImproved checksum verification with detailed warning messages

0.2.7Jan 18, 2026

Update Notifications

Automatic version update notifications at server startup with 24h local caching.

FeatureAutomatic version update notification at server startup

FeatureNon-blocking background check using GitHub Releases API

FeatureLocal cache (24h) to avoid repeated API calls

Feature--no-update-check CLI flag to disable update notifications

FeatureSEMANTIQ_UPDATE_CHECK environment variable for configuration

ImprovementUpdated author info to keyldzn

0.2.6Jan 18, 2026

Automatic Reindexation

Automatic reindexation when parser version changes, plus improved TypeScript/JavaScript support.

FeatureAutomatic reindexation when parser version changes (no more manual --force needed)

FeaturePARSER_VERSION constant to track parser logic changes

FeatureSupport for const/let variable extraction in TypeScript/JavaScript

FeatureGitHub Sponsors funding configuration

ImprovementVersion detection uses atomic transactions to prevent race conditions

ImprovementDocumentation updated with known limitations and setup guides

FixFilter out verbose ONNX Runtime logs during indexing

0.2.4Jan 18, 2026

ONNX Stability Fixes

Critical fixes for ONNX model download and inference, plus improved file exclusion handling.

FixModel download failing in async Tokio context (replaced reqwest::blocking with ureq)

FixDownload size limit too small for 90MB ONNX model (increased to 200MB)

FixONNX inference crash due to missing token_type_ids input

FixEmbeddings not generated during semantiq index command

Improvementsemantiq index now generates embeddings for all chunks

ImprovementCentralized file exclusion logic into exclusions.rs module

ImprovementAuto-indexer and FileWatcher now use shared exclusion patterns

0.2.3Jan 18, 2026

Semantic Search with ONNX

Semantic code search via ONNX embeddings, automatic model download, and cosine similarity matching.

FeatureONNX embedding model integration for semantic search

FeatureAutomatic model download on first run

FeatureCosine similarity search for vector matching

FeatureAlternative installation via cargo install --git

FeatureCHANGELOG.md for version history

ImprovementEmbeddings now generated automatically during auto-indexing

ImprovementSwitch from OpenSSL to rustls for better cross-compilation support

ImprovementUse ort download-binaries for automatic ONNX Runtime provisioning

BreakingRemoved macOS Intel (x86_64-apple-darwin) binary - ONNX Runtime does not support this target

0.2.2Jan 17, 2026

CLAUDE.md Improvements

Updated CLAUDE.md template to prioritize Semantiq MCP tools.

ImprovementImproved CLAUDE.md template to prioritize Semantiq tools over grep/Glob

0.2.1Jan 17, 2026

Performance & Security Fixes

Critical fixes for error handling, security, and query performance.

FixError handling with proper mutex propagation

FixSQL injection vulnerability via LIKE escaping

FixUTF-8 safety in tree-sitter text extraction

FixN+1 query pattern in get_stats() (4 queries → 1)

ImprovementShared single Arc<IndexStore> instead of 3 separate DB connections

ImprovementImproved scoring algorithm with symbol type boosting

ImprovementResults limited to 500 to prevent memory issues

ImprovementAdded PRAGMA busy_timeout=5000 for concurrent access

0.2.0Jan 17, 2026

NPM Package Automation

Automatic npm package version synchronization from git tags.

FeatureAutomatic npm package version update from git tag

0.1.3Jan 17, 2026

Project Init Command

New init command for project setup with automatic configuration.

FeatureNew semantiq init command for easy project setup

FeatureAuto-creates .claude/settings.json, CLAUDE.md, updates .gitignore

FeatureRuns initial indexation automatically

0.1.2Jan 17, 2026

Auto-Indexing

Real-time file updates with automatic re-indexing on changes.

FeatureAuto-indexing for real-time file updates

FeatureFileWatcher integration with create/modify/delete events

FeatureBackground task with 2-second polling

0.1.1Jan 17, 2026

Documentation Update

Improved npm and project documentation.

Featurenpm README documentation

ImprovementUpdated main README with correct npm package name

0.1.0Jan 17, 2026

Initial Release

First public release of Semantiq with core MCP functionality.

FeatureInitial release

FeatureMCP server with 4 tools: search, find_refs, deps, explain

FeatureSupport for 9 languages via tree-sitter

FeatureSQLite storage with FTS5 search