Recall Global — Retrieve Cross-Project Knowledge

SkillSearch

Search and retrieve global memories — knowledge that applies across all projects. Use when the user asks 'what are our coding standards', 'what conventions do we follow', 'what's our infrastructure setup', 'do we have a rule about', 'what applies to all projects', 'shared knowledge', 'global rules', or when you need cross-project context like architecture decisions, server configs, or team policies.

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the Recall Global — Retrieve Cross-Project Knowledge skill

What this skill tells your AI

The instructions your AI receives, as published by cdeust/cortex in skills/cortex-recall-global/SKILL.md and read by ahel’s review.

Keywords

global, convention, standard, rule, infrastructure, policy, all projects, shared, universal, what's our rule, coding standard, architecture rule, team agreement, cross-project, server config, deployment

Overview

Retrieve global memories — knowledge stored as cross-project that's visible regardless of which project you're currently working in. Global memories include architecture rules, coding conventions, infrastructure facts, security policies, and team agreements.

Note: Regular cortex:recall already surfaces global memories automatically. This skill is for when you specifically want to focus on cross-project knowledge.

Workflow

Step 1: Recall Global Knowledge

Query with any domain — global memories appear alongside domain-specific results:

cortex:recall({
  "query": "<topic to search for>",
  "max_results": 10
})

Global memories are included in results regardless of the current project domain.

Step 2: Filter to Global Only

To see only global cross-project knowledge, use the unified neural graph:

cortex:open_visualization()

Click the Global filter button (pink) to isolate all global memories.

Click the Global filter — global memories appear as pink nodes connected to all project domains.

Step 3: Explore by Category

Common global recall patterns:

Architecture rules:

cortex:recall({ "query": "architecture rules and principles" })

Infrastructure:

cortex:recall({ "query": "server addresses and database connections" })

Coding conventions:

cortex:recall({ "query": "coding standards and naming conventions" })

Security policies:

cortex:recall({ "query": "security policies and credential management" })

Step 4: Navigate Connections

After finding a global memory, explore what it connects to across projects:

cortex:navigate_memory({
  "memory_id": <id>,
  "depth": 2
})

Global memories link to all domain hubs in the knowledge graph — following connections shows which projects reference similar concepts.

How Global Recall Works

Global memories have is_global = TRUE in the database. During recall, every retrieval signal (vector, FTS, trigram, heat, recency) includes the clause:

WHERE (domain = current_domain OR is_global = TRUE)

This means global memories compete on relevance alongside domain-specific ones — they're not artificially boosted, just not filtered out.

Tips

  • Global memories compete on merit: They appear in results only when relevant to the query, not automatically at the top
  • Use the visualization: The unified graph shows global memories (pink) with edges to every project — a visual map of shared knowledge
  • Rate for quality: Use cortex:rate_memory on global memories that were helpful to improve future retrieval confidence
  • Assess coverage: Use cortex:assess_coverage to see if any project domain is missing shared knowledge

Signals

GitHub stars
72
Forks
11
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
cortex-recall-global
Source
github.com/cdeust/cortex