nebula-memory

Rules for maintaining a shared work log in a coding project. The log has an index, a list of lasting warnings, and detailed entries that agents can consult before and after tasks.

In plain words
What is it for?
Use it before a task to find relevant decisions or warnings, and after code changes, bug diagnosis, or non-obvious discoveries to record the result.
Why use it?
It gives agents prior project context and recorded pitfalls, reducing repeated work and avoidable mistakes. It also says when entries should be added and which files to update.

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/agentsystemlabs/nebula/nebula-memory
Clone the repo
git clone --depth 1 https://github.com/AgentSystemLabs/nebula

Made for: Cursor.

Per session 350 This file is loaded in full into every session.
When invoked 350 The same file — it is already loaded in full.
Security scan B 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 $0.00350 $0.00350
Opus 5 $0.00175 $0.00175
Sonnet 5 $0.00070 $0.00070
Haiku 4.5 $0.00035 $0.00035

Measured 2d ago against content hash 3fbba7580191, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

nebula-memory scanned grade B 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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

(`grep -ril '<symbol or TERM>' .claude/memory/entries` when the index is not enough). A recorded gotcha
.cursor/rules/nebula-memory.mdc · 22 lines

What it actually says

This repo keeps a shared MEMORY LOG in three layers: .claude/MEMORY.md is the index (one line per task — date, title, the TERMS and files it is about — newest first), .claude/memory/gotchas.md holds the standing gotchas (the traps that outlive their task, one line each, grouped by TERM), and .claude/memory/entries/<date>-<slug>.md hold each task's full Asked / Did / Gotchas.

Before starting a task, read .claude/MEMORY.md and .claude/memory/gotchas.md in full, then open the entry files whose index line names the TERMS, crate, symptom or file the user is asking about (grep -ril '<symbol or TERM>' .claude/memory/entries when the index is not enough). A recorded gotcha is a mine already stepped on; a recorded decision is settled unless the user reopens it. Entries describe what was true when written, so confirm any file or flag they name still exists before relying on it.

After finishing a task that changed code or behavior, diagnosed a bug, or turned up something non-obvious, read .claude/skills/nebula-memory/SKILL.md and follow it: one entry file, one index line, and any durable trap into the standing gotchas. Skip it for pure questions and trivial edits — the log is only useful if it stays free of restated diffs. Run make memory-check afterwards.

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 · 22 lines · 350 tokens per session scan B 3fbba7580191

Subscribe to this mod's changes

nebula-memory is a cursor rule published in the GitHub repository AgentSystemLabs/nebula (100 stars, last pushed yesterday), licensed MIT. It adds 350 tokens to every session, about $0.0018 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.