BRUTHA RELEASE NOTES & CHANGELOG

What shipped in every release.

Review the production v1.0.0 release notes for Brutha Solo and Brutha Enterprise. Future release updates will appear here.

SOLO EDITION // RELEASE NOTES

Brutha Solo Releases

Production release notes for Brutha Solo. Every capability below runs locally on your machine with zero cloud egress.

v1.0.0

Latest Release
2026-10-21

Production release of Brutha Solo

Brutha Solo v1.0.0 packages local persistent memory, portable Characters, the Persistent Code Knowledge Graph, automatic duplicate and contradiction checking, and one-command client setup into a single self-contained binary.

Linux x86_64 (glibc 2.38+)Linux ARM64macOS Apple Silicon (ARM64)Windows x86_64

Key Highlights

  • 32 MCP tools for persistent memory, Characters, and codebase intelligence.
  • 100% local runtime with SQLite by default and single-owner local Postgres support.
  • Automatic duplicate corroboration and contradiction warnings when saving new facts.
  • Portable Characters that carry your working instructions and preferred model hints across AI tools.
  • Persistent Code Knowledge Graph with uncommitted working-tree overlays and blast-radius analysis.

Persistent Memory & Retrieval

  • Two-tier retrieval returns compact summaries first (`memory_recall`, `memory_search`) and loads full text only on demand (`memory_expand`), cutting context token usage while keeping full recall accuracy.
  • Automatic write-time duplicate and contradiction checking strengthens existing memories when you save a matching fact and flags conflicts so your assistant can supersede outdated decisions cleanly.
  • Temporal validity windows (`valid_from` and `valid_until`) are enforced consistently across `memory_recall`, `memory_search`, linked context expansions, and `memory_synthesize` dossiers.
  • Scheduled memory decay (`memory_decay`) automatically demotes un-recalled memories from hot to warm to cold without deleting data, keeping active context sharp.
  • Live token-savings reporting (`memory_token_savings`) tracks characters and tokens saved across your current session, week, month, or all time.

Portable Characters & Session Lifecycle

  • Create, activate, and incrementally patch reusable personas with `character_save`, `character_update`, `character_activate`, `character_list`, `character_release`, and `character_archive`.
  • Characters support an optional preferred model hint and workspace default primer binding so every new session starts with your role rules and project context.
  • SessionWatch (`end_session_review`) tracks session activity and prompts your assistant to capture key decisions and outcomes before a conversation closes.

Persistent Code Knowledge Graph (13 Tools)

  • Index repositories and map architecture with `code_map`, `code_index_repo`, `code_search`, `code_symbol`, `code_neighbors`, `code_trace`, `code_impact`, and `code_architecture`.
  • Local working-tree overlay automatically merges uncommitted file edits over the latest indexed commit snapshot so impact queries reflect code you just wrote.
  • Link decisions and runbooks directly to code symbols (`code_memory_link`, `code_memory_context`) and track symbol drift (`code_subscribe`, `code_outbox_claim`, `code_proposal_resolve`).
  • Cross-repository dependency linking connects package manifests and service routes across repositories indexed in the same store.

Client Integrations & Operations

  • One-command setup (`brutha-mcp install`) auto-configures Claude Code, OpenAI Codex CLI, Gemini CLI, Cursor, Windsurf, Cline, and Continue.dev.
  • Local HTTP daemon (`brutha-mcp http` and `brutha-mcp daemon install`) exposes memory and Character endpoints for custom scripts and non-MCP agent frameworks.
  • Built-in self-updater (`brutha-mcp update`), Markdown vault importer (`brutha-mcp import-docs`), and daily-rotated logs (`~/.brutha/logs/brutha.log`).