remind AGENTS.md

A development guide for Remind, a memory system that stores facts and relationships for AI agents. It covers the project’s code structure, data models, storage, retrieval, and configuration.

In plain words
What is it for?
Use it when developing Remind itself, including its memory models, database layer, command-line interface, retrieval, and fact-processing code.
Why use it?
It gives coding agents the background needed to change Remind without guessing how its parts fit together.

Instructions file for CodexOpenCode

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/sandst1/remind/agents-md
Clone the repo
git clone --depth 1 https://github.com/sandst1/remind

Made for: Codex, OpenCode.

Per session 3,149 This file is loaded in full into every session.
When invoked 3,149 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.03149 $0.03149
Opus 5 $0.01574 $0.01574
Sonnet 5 $0.00630 $0.00630
Haiku 4.5 $0.00315 $0.00315

Measured yesterday against content hash 44626315c240, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

remind AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

AGENTS.md · 318 lines

How it starts

The opening of the file, as written. The whole thing — 318 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Remind - Development Guide for AI Agents

This guide is for AI agents developing Remind itself. For using Remind as a memory layer, see docs/AGENTS.md.

Project Overview

Remind is an agent-driven memory layer for LLMs. It provides temporal facts, semantic retrieval, and structured curation — the calling agent is the only intelligence. There are no internal LLM calls.

Core architecture: Episodes → Agent curation via apply → Concepts with relations

Architecture

src/remind/
├── models.py          # Data models (Concept, Episode, Entity, Relation, Fact, Conflict)
├── store.py           # SQLAlchemy persistence layer (SQLite, PostgreSQL, MySQL)
├── interface.py       # MemoryInterface - main public API
├── config.py          # Configuration loading (config file, env vars, defaults)
├── facts.py           # Deterministic fact processing (clustering, collision detection)
├── apply.py           # Batch write engine (op vocabulary, transaction support)
├── snapshot.py        # Batch read engine (combinable scopes)
├── retrieval.py       # Spreading activation retrieval
├── reranker.py        # Optional cross-encoder reranking (requires [rerank] extra)
├── cli.py             # Command-line interface (project-aware)
├── mcp_server.py      # MCP (Model Context Protocol) server
├── background.py      # Background recall worker spawning
├── background_worker.py # Subprocess entry point for recall worker
├── api/               # REST API for web UI
│   ├── __init__.py    # Exports api_routes
│   └── routes.py      # Starlette route handlers
├── static/            # Web UI assets (compiled)
│   ├── index.html     # Entry point
│   └── assets/        # CSS/JS bundles
├── skills/            # Bundled agent skills (installed via `remind skill-install`)
│   ├── remind-capture/  # When/how to write memories while working
│   ├── remind-context/  # When/how to recall before acting
│   └── remind-curate/   # Consolidation procedure, conflict triage, label upkeep
└── providers/         # Embedding provider implementations
    ├── base.py        # EmbeddingProvider ABC
    ├── local.py       # Local embeddings via fastembed (default)
    ├── openai.py      # OpenAI embeddings
    ├── azure_openai.py # Azure OpenAI embeddings
    └── ollama.py      # Ollama embeddings

Read the full file on GitHub · 318 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 318 lines · 3,149 tokens per session scan A 44626315c240

Subscribe to this mod's changes

remind AGENTS.md is an instructions file published in the GitHub repository sandst1/remind (83 stars, last pushed 22d ago), licensed Apache-2.0. It adds 3,149 tokens to every session, about $0.0157 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.