ai-memory

A set of rules for using Ratary as an organisational memory service while treating the docs-ai repository as the authoritative knowledge base. Ratary helps an agent recall linked project information, while the code repository remains the authority for implementation details.

In plain words
What is it for?
Use it during Ontorata development sessions for recovering context, checking decisions against project files, saving handoff notes, and keeping organisational records synchronised.
Why use it?
It prevents remembered context from replacing current project documents or code. It also defines what to check at the start and end of a session and what to do when Ratary cannot be reached.

Cursor rule for Cursor

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 rules/ontorata/ratary/ai-memory
Clone the repo
git clone --depth 1 https://github.com/ontorata/ratary

Made for: Cursor.

Per session 447 This file is loaded in full into every session.
When invoked 447 The same file — it is already loaded in full.
Security scan A 1 finding. 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 $0.00447 $0.00447
Opus 5 $0.00224 $0.00224
Sonnet 5 $0.00089 $0.00089
Haiku 4.5 $0.00045 $0.00045

Measured 2d ago against content hash a7a325b4fa99, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ai-memory scanned grade A with 1 finding 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 2d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- **Pakai MCP tools** — bukan REST curl untuk memory
.cursor/rules/ai-memory.mdc · 45 lines

What it actually says

Ratary — Organizational Memory

Arsitektur: docs-ai/cross-cutting/ratary-sync · docs-ai/INDEX.md

Three layers

Layer Role
docs-ai Knowledge OS Source of truth — phases · ADR · decisions (Git)
Ratary (MCP) Semantic organizational memory — active recall · links to docs-ai
Cursor Execution — implement · sync docs-ai · validate

Session start (when MCP available)

Recover Ontorata context using Ratary first. Validate against docs-ai Knowledge OS.

  1. search_memory — project ontorata / ratary · tags handoff
  2. get_memory_by_codename if user mentions codename
  3. Validate against docs-ai files and product code — docs-ai wins on knowledge; code repo wins on implementation
  4. Fallback if MCP down: sessions/CURRENT.md

Session end

  1. save_memory — primary handoff (tags: handoff, project name) · include docs-ai/... path when relevant
  2. Update docs-ai session audit if needed
  3. Sync config lives under docs-ai cross-cutting/ratary-sync/

Rules

  • Jangan anggap Ratary menggantikan file di docs-ai
  • Jangan minta API key di Cursor — MCP sudah akses D1
  • Pakai MCP tools — bukan REST curl untuk memory
  • Bootstrap laptop baru: clone product repos + docs-ai → env → baru Ratary

Phase 4

Dogfood: Ontorata organizational memory = first internal production workload on Ratary.

Balas user dalam Indonesian; memory content / ADR dalam English.

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. 2d ago First seen · 45 lines · 447 tokens per session scan A a7a325b4fa99

Subscribe to this mod's changes

ai-memory is a cursor rule published in the GitHub repository ontorata/ratary (1 stars, last pushed 23d ago), licensed MIT. It adds 447 tokens to every session, about $0.0022 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.