GenAIScope v0.4.0

GenAIScope v0.4.0 - Universal Memory Access

Local-first AI memory now speaks every protocol: MCP, REST, OpenAI, Anthropic, Gemini, and vector search - all zero-dependency by default.

---

What's new

Semantic & Hybrid Search

• Three search modes: keyword (v0.3 behaviour), vector (cosine similarity), hybrid (fused score - default)
• Pluggable embedding backends: LocalHashEmbedder (zero-dep, pure Python), SentenceTransformerEmbedder, OpenAIEmbedder
• LocalVectorStore (SQLite-backed) and RedisVectorStore - drop-in swappable
• MemorySearchResult now carries vector_score, keyword_score, fused_score, embedder_name
• Semantic cache upgraded to embedding cosine similarity (threshold 0.92, deterministic fallback retained)

memory.context() - Injectable Context Block

ctx = store.context("coding style preferences", max_chars=2000)
system_prompt = ctx.text  # ready to inject into any LLM call

GenAIScope v0.3.0 - Production Memory Backend for AI Apps and Memovo

Full Changelog: v0.2.91...v0.3.0

GenAIScope v0.3.0 Release Notes

GenAIScope v0.3.0 is the production memory backend release.

Highlights

• Pluggable SQLite and optional Redis memory backends
• User, workspace, project, agent, and session scoped memory
• TTL cleanup, deterministic hybrid search, dedupe, and JSON/JSONL migration
• Redis trace logging and lightweight semantic cache foundation
• Backend-aware dashboard reporting and Memovo-ready APIs

Known Limitations

• Real embedding and Redis vector search support is planned for v0.4.0
• MCP server support is planned for v0.5.0
• Qdrant and pgvector integrations are planned for later releases

GenAIScope v0.2.91

GenAIScope v0.2.91 Release Notes

GenAIScope v0.2.91 adds a local-first memory, file intelligence, prompt coaching,
trace logging, and dashboard layer on top of the existing readiness toolkit.

Install

pip install genaiscope

New Features

• SQLite-backed local memory store
• Local keyword/hybrid memory search
• Prompt quality coach
• File memory for TXT, MD, JSON, and CSV
• Local trace logging
• Static HTML dashboard

CLI

genaiscope memory add "User prefers concise answers" --type preference
genaiscope memory add-prompt "Summarize this properly."
genaiscope memory search "concise answers"
genaiscope files add README.md
genaiscope trace stats
genaiscope dashboard generate

Python

from genaiscope.memory import MemoryStore

memory = MemoryStore()
memory.add("User prefers short CTO-level answers.", memory_type="preference")
print(memory.search("answer style"))

Known Limitations

• SQLite only in this release
• Local keyword/hybrid scoring only; no embeddings yet
• PDF and DOCX ingestion are not included
• Dashboard is static HTML

Roadmap

• Redis backend
• Vector search
• Semantic cache
• MCP memory server
• REST API

GenAIScope v0.1.0

GenAIScope v0.1.0 - Initial Alpha Release

GenAIScope is now live as an open-source Python package and CLI toolkit for inspecting, testing, securing, optimizing, and operationalizing GenAI applications before production.

This first release establishes the foundation for GenAIScope as a lightweight, local-first, developer-friendly toolkit to help teams detect what is risky, broken, expensive, or unreliable in GenAI applications before users do.

---

Installation

pip install genaiscope