惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

量子位
Recent Announcements
Recent Announcements
D
Docker
V
V2EX
阮一峰的网络日志
阮一峰的网络日志
Vercel News
Vercel News
Microsoft Security Blog
Microsoft Security Blog
The GitHub Blog
The GitHub Blog
U
Unit 42
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
月光博客
月光博客
腾讯CDC
B
Blog
博客园_首页
罗磊的独立博客
D
DataBreaches.Net
IT之家
IT之家
酷 壳 – CoolShell
酷 壳 – CoolShell
L
LangChain Blog
aimingoo的专栏
aimingoo的专栏
MongoDB | Blog
MongoDB | Blog
GbyAI
GbyAI
Stack Overflow Blog
Stack Overflow Blog
M
MIT News - Artificial intelligence

Hacker News: Show HN

PurrrrrFocus: Pomodoro Timer App - App Store Workflow Engine — Multi-Step Orchestration for Bun RapidPhoto: Pro Photo Editor App - App Store GitHub - DheerG/swarms: Achieve extraordinary results with claude code across a variety of tasks SPICE simulation → oscilloscope → verification with Claude Code — Lucas Gerads Show HN: VCoding – A 5 MB native Windows IDE with no dynamic dependencies Show HN: LLMs don't hallucinate because they're bad at math, it's the format GitHub - Agent-FM/agentfm-core: AgentFM is a peer-to-peer network that turns everyday computers into a decentralized AI supercomputer. AgentFM lets you run massive AI workloads directly across a global mesh of idle CPUs and GPUs. Show HN: Tracking Top US Science Olympiad Alumni over Last 25 Years GitHub - Potarix/agent-hub: One place to talk to all your agents Show HN: Runtime security for AI agents(injection,tool abuse, data exfiltration) GitHub - dubeyKartikay/lazyspotify: Terminal Spotify client for macOS and Linux GitHub - the-banana-tool/king-louie: Easy to use GUI Personal AI Assistant. Win/Linux/Mac. Show HN I made my vacation rental bookable by AI agents–no Airbnb, 0% commission GitHub - basteez/jsf-autoreload: maven plugin to enable hot reload on jsf projects uvm32/hosts/host-gdbstub at main · ringtailsoftware/uvm32 GitHub - labsai/EDDI: Config-driven engine that turns JSON into production-grade AI agents. Multi-agent orchestration, 12+ LLM providers, MCP/A2A protocols, RAG, persistent memory, and enterprise compliance (EU AI Act, GDPR, HIPAA). Built on Quarkus. GitHub - glitchnsec/fortyone-oss: AI Executive Assistant Platform Quickstart | Alien GitHub - muxshed/shed: One stream in, or many. Every destination, simultaneously. No cloud middleman, no per-channel fees, no limits. GitHub - ocrbase-hq/ocrbase: 📄 PDF/IMG ->.MD/JSON Document OCR API for PaddleOCR and GLMOCR. Self-hostable. GitHub - impactjo/home-memory: MCP server that lets your AI assistant remember everything about your home. GitHub - Sets88/dbcls: DbCls is a powerful terminal database client that supports various databases GitHub - neptun2000/heor-agent-mcp GitHub - SeanFDZ/macmind: Single-layer transformer in HyperTalk for the classic Macintosh RollQuation: Math Puzzles - Apps on Google Play GitHub - dropbox/witchcraft Show HN: Agent-cache – Multi-tier LLM/tool/session caching for Valkey and Redis GitHub - opentalon/opentalon: OpenTalon is an open-source platform built from the ground up in Go as a robust alternative to OpenClaw LinkedIn™ 职位抓取工具 - Chrome 应用商店
GitHub - Chadi00/thr: local semantic memory for your term...
chadiiek · 2026-05-04 · via Hacker News: Show HN

Local semantic memory for your terminal and coding agents.

Latest release Platform License

thr is a small local CLI for saving explicit memories and recalling them later by meaning or text. It is built for agent workflows: stable JSON output, offline semantic search, and installable skills for Codex, OpenCode, and Claude Code.

thr semantic memory demo

Install

curl -fsSL https://raw.githubusercontent.com/Chadi00/thr/master/install.sh | bash

The installer downloads the latest macOS or Linux release and verifies signed checksums before installing. Prefer manual verification? See MANUAL_INSTALL.md.

Quick Start

thr add "This project prefers small PRs with tests"
thr ask "how should I structure this PR?"
thr search "small PRs"
thr list

thr ask performs semantic recall over your saved memories. It does not call an LLM or generate an answer.

Agents

Install the thr skill for your coding agent:

thr setup codex
thr setup opencode
thr setup claude-code

After that, an agent can retrieve durable project facts with thr ask --json or thr search --json, then maintain memories with thr add, thr edit, and thr forget.

Features

  • Local first: memories live in your own SQLite database.
  • Semantic and text recall: use ask for meaning, search for exact or fuzzy text.
  • Agent ready: JSON output and packaged skills make it easy for agents to use safely.
  • Offline by default: the embedding model and ONNX Runtime are bundled with the release.
  • Explicit memory: thr stores only what you save.

CLI

thr add <text>          Save a memory
thr add -               Save stdin
thr list                List memories and ids
thr show <id>           Print one memory
thr ask <question>      Retrieve semantically similar memories
thr search <query>      Search memories by text
thr edit <id> <text>    Replace a memory
thr forget <id>         Delete a memory
thr stats               Show database path and count
thr index               Rebuild semantic embeddings
thr prefetch            Prepare the model cache
thr setup <agent>       Install an agent skill

Useful flags:

thr --db ./thr.db ...
THR_DB=./thr.db thr list
thr ask --json "repo conventions"
thr search -n 5 "release notes"

Data

Data Default
Memories ~/.thr/thr.db
Model cache ~/.thr/models
Install prefix ~/.local

Memories are stored as local plaintext SQLite. The default data directories are created with private filesystem permissions, but the database is not encrypted at rest.

thr has no telemetry. Release archives include the embedding model and runtime needed for offline semantic search.

Platform

Prebuilt releases support macOS arm64/x86_64 and glibc Linux arm64/x86_64.

Windows and Alpine/musl Linux are not supported yet.

Links