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.
npx agentmods add instructions/memo-fs/memofs/agents-mdgit clone --depth 1 https://github.com/memo-fs/memofsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/memo-fs/memofs/agents-md)<a href="https://agentmods.dev/instructions/memo-fs/memofs/agents-md"><img src="https://agentmods.dev/badge/instructions/memo-fs/memofs/agents-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00646 | $0.00646 |
| Opus 5 | $0.00323 | $0.00323 |
| Sonnet 5 | $0.00129 | $0.00129 |
| Haiku 4.5 | $0.00065 | $0.00065 |
Grade A, and why
memofs 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 3d ago.
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.
How it starts
The opening of the file, as written. The whole thing — 42 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MemoFS — Agent Guidelines
MemoFS is the open-source, file-first memory runtime for AI agents.
Monorepo Architecture
- Package Manager & Monorepo:
pnpmworkspace with Turborepo. - Package Scopes:
@memofs/*: Public, published packages (packages/*).@repo/*: Private internal tooling and build configs (tooling/*).
- Core Runtime & Entrypoints (
@memofs/core):- Universal/Edge entrypoint:
@memofs/core(no Node.js-specific dependencies). - Node.js filesystem entrypoint:
@memofs/core/node-fs. - Cloud replication client:
@memofs/core/cloud-client.
- Universal/Edge entrypoint:
- Tooling Stack: TypeScript, Biome for formatting and linting, tsdown / rolldown for builds, Vitest for testing.
Core Behavioral Principles
- Positioning & Terminology: MemoFS is a file-first memory runtime for AI agents in general (not exclusively coding agents). Always balance framing 50/50 between users of AI agents (running agents day-to-day via MCP or lifecycle hooks) and builders of AI agents (building agents and frameworks with
@memofs/coreor@memofs/server). - Facts Over Assumptions: Always inspect source code and verify contracts directly. Do not rely on unverified assumptions.
- Package Boundaries: Strictly preserve package isolation. Do not import private
@repo/*packages into public@memofs/*distributions. - Code Style & Modularity: Keep source and test files cohesive and modular (target
<= 500 LOCper file). Use structured logging rather than rawconsole.login library code. - Build Configurations: Reuse shared build configs from
@repo/tsdown(pkgConfig) rather than duplicating build flags. - Clean Git Hygiene: Never commit secrets, API keys,
.envfiles, or gitignored runtime data.
Workspace Rules
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.
- 3d ago First seen · 42 lines · 646 tokens per session scan A ad16a8ecf330
memofs AGENTS.md is an instructions file published in the GitHub repository memo-fs/memofs (6 stars, last pushed 3d ago), licensed MIT. It adds 646 tokens to every session, about $0.0032 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 2026-08-31.
Other instructions, from other repositories
zer0dex AGENTS.md
Instructions for hermes-labs-ai/zer0dex, covering agents.md, use it for, do not use it for, minimal commands and output shape.
engram CLAUDE.md
Instructions for techtheist/engram, covering engram — context for claude code, hard rules (locked — don't relitigate without reason), where things go, workflows & sharp edges and now / next (keywords).
vexic AGENTS.md
Instructions for PyroDonkey/vexic, covering vexic agent instructions, project, architecture boundaries, package boundary and host ports.
codemem CLAUDE.md
Instructions for kunickiaj/codemem, a project described as: A lightweight persistent-memory companion for OpenCode & Claude.
ai-memory CLAUDE.md
Claude Code instructions for blockmandev/ai-memory, covering project memory instructions, when to save memories (addmemory), memory types, importance levels and when to search memories (searchmemories).
hindsight CLAUDE.md
Claude Code instructions for vectorize-io/hindsight, covering claude.md, project overview, development commands, local development (api + ui) and start both api server and control plane ui.