v0.1.3

Fix CI: add explicit toolchain: stable for pinned rust-toolchain

The pinned dtolnay/rust-toolchain v1 commit enforces required: true
on the toolchain input — unlike the @stable tag which implied it.
Add toolchain: stable to both CI and release workflows.

Also pick up cargo fmt formatting fix in runtime.rs.

v0.1.2

CogZ v0.1.2

Patch release fixing broken install URLs.

Fixes

• Install URLs pointed to wrong branch — the README install one-liners and install script comments referenced main, but the repository's default branch is master. All four URLs would 404, making the documented install commands non-functional for new users. Fixed main → master in README.md, install.sh, and install.ps1.

Upgrade

cogz update

Or install fresh:

curl -fsSL https://raw.githubusercontent.com/balaianu/CogZ/master/install.sh | bash

Checksums

See SHA256SUMS in the release assets for binary integrity verification.

v0.1.1

CogZ v0.1.1

Bug-fix release addressing two issues that blocked clean installs.

Fixes

• Install script JSON parsing — the install.sh grep pattern for extracting browser_download_url from the GitHub API response did not match the actual JSON key format, producing a malformed URL and a curl exit 3. The install script was broken for all users. Fixed the regex to match "browser_download_url": "...".
• ONNX Runtime auto-download checksum handling — the ORT auto-download treated a missing SHA256SUMS manifest (404) as a hard failure, but Microsoft ORT releases don't ship SHA256SUMS files. This made auto-download always fail on clean installs, preventing embeddings from working. Per the security policy, a missing manifest is now degraded mode (HTTPS-only download with a warning). The manifest-present-but-asset-missing case still fails hard — that's the supply-chain red flag.

Documentation

• Agent compatibility — README and agent-setup.md updated. Removed Windsurf (acquired by Devin). Added Gemini CLI and GitHub Copilot with config paths and docs links. Agent list now leads with Claude Code and Cursor instead of Devin.
• Docs cleanup — removed Python predecessor comparison from architecture.md, deleted stale hot-path evaluation, removed version stamp from resource profile, restructured agent-setup.md into equal per-agent sections with shared hook config centralized in hooks.md.
• Release workflow — added permissions: contents: write to fix the HTTP 403 release creation error.

Upgrade

cogz update

Or install fresh:

curl -fsSL https://raw.githubusercontent.com/balaianu/CogZ/master/install.sh | bash

Checksums

See SHA256SUMS in the release assets for binary integrity verification.

v0.1.0

CogZ v0.1.0

CogZ is a local-first, code-aware engineering cognition runtime for AI coding agents. It gives a coding agent persistent memory, contextual retrieval, and continuous cognition about a software repository — all running locally, no cloud services required.

Features

• Persistent memory — observations, rules, and knowledge stored as Markdown files, structured by taxonomy and linked to the code itself
• Code-aware search — hybrid FTS5 + vector search with graph expansion, separate embedding models for code and knowledge
• Context packs — scoped, ranked, token-budgeted context with graph provenance, assembled in three modes (cold start, task, escalation)
• Code indexing — tree-sitter parses Rust, Python, Go, JavaScript, TypeScript, TSX, and Bash into a graph of functions, classes, files, and modules with structural edges
• Consolidation — dedup detection, NLI-based contradiction flagging, observation-to-rule promotion, duplicate merging
• Lifecycle hooks — session start, prompt submit, file save, session end — injects context packs and triggers incremental reindexing
• Graceful degradation — works fully without ONNX models in FTS-only mode; models auto-download on first use and auto-unload when idle
• Agent-agnostic — MCP server (13 tools) + shell hooks, works with Devin, Claude Code, Cursor, Codex, Windsurf, and any MCP-compatible agent

Install

curl -fsSL https://raw.githubusercontent.com/balaianu/CogZ/master/install.sh | bash

Quick start

cd ~/your-project
cogz init
cogz index
cogz search "authentication flow"

Requirements

• Linux x86_64/aarch64, macOS arm64, or Windows x86_64
• ~550 MB disk (binary + ONNX Runtime + 3 models)
• ~11 MB RAM idle, ~300-500 MB with models loaded

Documentation

• Getting Started
• Configuration
• CLI Reference
• MCP Tools
• Agent Setup
• Architecture

Checksums

See SHA256SUMS in the release assets for binary integrity verification.