Skip to content

/cortex-remember-global

Store a global memory that is visible across all projects. Use when the user shares architecture rules, coding conventions, infrastructure facts, security policies, team agreements, or any knowledge that applies beyond a single project. Triggers on 'remember this everywhere',

From plugin
6813 skills3 commands
shell
$ npx -y skills add cdeust/Cortex --skill cortex-remember-global --agent claude-code

How it fires

How this skill gets triggered: by you, by Claude, or both.

  • Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.
  • You can call itInvoke it directly when you want it.
  • Slash command/cortex-remember-global
How auto-invocation works

Context preview

The summary Claude sees to decide when to auto-load this skill.

Store a global memory that is visible across all projects. Use when the user shares architecture rules, coding conventions, infrastructure facts, security policies, team agreements, or any knowledge that applies beyond a single project. Triggers on 'remember this everywhere',

SKILL.md

cortex-remember-global.SKILL.md
name: cortex-remember-global
description: "Store a global memory that is visible across all projects. Use when the user shares architecture rules, coding conventions, infrastructure facts, security policies, team agreements, or any knowledge that applies beyond a single project. Triggers on 'remember this everywhere', 'this applies to all projects', 'global rule', 'shared convention', 'infrastructure note', 'cross-project', or when the content is clearly universal (clean architecture, SOLID, deployment configs, server addresses)."

Remember Global — Store Cross-Project Knowledge

Keywords

global, everywhere, all projects, cross-project, shared, universal, convention, standard, rule, infrastructure, policy, team agreement, always, never, architecture rule, coding standard

Overview

Store knowledge that transcends any single project into Cortex's global memory. Global memories bypass domain filtering during recall — they're visible from every project you work on. Use this for architecture rules, coding conventions, infrastructure facts, security policies, and team agreements.

**Cortex auto-detects many global patterns** (clean architecture, dependency injection, server addresses, etc.), but use this skill explicitly when you want to guarantee cross-project visibility.

Workflow

Step 1: Identify Cross-Project Knowledge

Good candidates for global memory:

  • **Architecture rules**: "Always follow clean architecture — inner layers never import outer layers"
  • **Coding conventions**: "Use UTC timestamps in all database layers"
  • **Infrastructure**: "Production database at db.internal:5432, daily backups at 3AM UTC"
  • **Security policies**: "Rotate API keys every 90 days, store in 1Password vault"
  • **Team agreements**: "PRs must be under 300 lines, always include tests"
  • **Reusable patterns**: "Use factory injection for all handler composition roots"

**Not global** (project-specific): bug fixes, feature decisions for one project, file-specific notes.

Step 2: Store as Global

cortex:remember({
  "content": "<clear, self-contained knowledge that applies across projects>",
  "tags": ["<category>", "<topic>"],
  "is_global": true,
  "force": true
})

**Content guidelines:**

  • Write as a rule or fact, not a narrative: "Always use UTC" not "Today we decided to use UTC in the auth service"
  • Include the *why* when it's a rule: "Use dependency injection because it enables testing and follows SOLID"
  • Keep it universal — no project-specific file paths, PR numbers, or branch names

Step 3: Verify Global Status

The response includes:

  • `is_global: true` — confirms cross-project visibility
  • `global_reason: "explicit"` — stored because you explicitly requested it

Step 4: Anchor for Permanence (Optional)

Global memories are already high-value, but if they must never decay:

cortex:anchor({
  "memory_id": <id>,
  "reason": "Core architecture rule — permanent"
})

Auto-Detection

Even without `is_global: true`, Cortex automatically detects global content using a weighted signal classifier across 6 categories:

| Category | Example signals | |---|---| | **Architecture** | clean architecture, SOLID, dependency injection, composition root | | **Convention** | coding standard, naming convention, best practice, team agreement | | **Infrastructure** | server at, database URL, Docker compose, CI/CD pipeline | | **Security** | API key rotation, credential policy, authentication, encryption | | **Cross-project** | all projects, shared across, universal, applies everywhere | | **Knowledge** | UTC timestamps, WAL mode, connection pools, idempotency |

If the weighted score exceeds threshold 3.0, the memory is automatically global — no explicit flag needed.

Tips

  • **Be declarative**: "Inner layers never import outer layers" is better than "We should probably avoid importing infrastructure in core"
  • **One rule per memory**: Don't bundle 5 conventions into one memory — store each separately for better retrieval
  • **Tag consistently**: Use `architecture`, `convention`, `infrastructure`, `security`, `policy` tags for easy filtering
  • **Review with visualization**: Use `cortex:open_visualization` and click the "Global" filter to see all cross-project knowledge
Read more
Read it on GitHub ↗
Ships withhypermnesia-mcp

Cross-platform persistent memory MCP for Codex, Gemini CLI, Claude Code, and other local MCP hosts. 36 cited neuroscience mechanisms, local-first SQLite/PostgreSQL, hybrid retrieval, decay-based consolidation, and reproducible benchmarks. Claude adds optional automatic lifecycle hooks.

Get the whole plugin, auto-invoked

Other skills on hypermnesia-mcp.