/cortex-profile

name: cortex-profile
description: "View and manage your cognitive profile — how you think, work patterns, blind spots, and cross-domain connections. Use when the user says 'show my profile', 'how do I work', 'what are my patterns', 'cognitive style', 'blind spots', 'methodology', or at the start of a session to load context. Also use 'rebuild profile' to rescan all session history, or 'list domains' to see all tracked project domains."

Profile — Cognitive Profiling and Methodology

Keywords

profile, methodology, cognitive style, work patterns, blind spots, how do I work, thinking style, domains, rebuild profile, my patterns, session history, cognitive, behavioral, persona

Overview

Cortex builds a cognitive profile from your Claude Code session history — how you explore code, which tools you prefer, what patterns you repeat, where you have blind spots, and how your style connects across different project domains. This profile is used to calibrate responses and surface relevant context.

**Use this skill when:** Starting a session (auto-injected via hook), wanting to understand your work patterns, or needing to rebuild profiles after significant new work.

Workflow

Step 1: Load Current Profile

cortex:query_methodology({
  "cwd": "<current working directory>",
  "first_message": "<what the user is working on>"
})

Returns:

• **Cognitive style** — Felder-Silverman dimensions (active/reflective, sensing/intuitive, visual/verbal, sequential/global)
• **Entry patterns** — How you typically start work in this domain
• **Recurring patterns** — Behavioral patterns that repeat across sessions
• **Blind spots** — Categories, tools, or patterns you tend to miss
• **Connection bridges** — Cross-domain analogies and structural similarities
• **Hot memories** — Most relevant active memories for this context
• **Fired triggers** — Prospective memories activated by the current context

Step 2: Explore Behavioral Features

Dive deeper into your behavioral profile:

cortex:explore_features({ "mode": "persona" })

Returns the 9-dimensional persona vector of every domain and their session-weighted composition (add `"domain"` for one domain): activeReflective, sensingIntuitive, sequentialGlobal, thoroughness, autonomy, verbosity, riskTolerance, focusScope, iterationSpeed. The response's `dimensions` lists them in vector order.

Other exploration modes:

• `"mode": "features"` — The sparse behavioral feature dictionary and the features that persist across domains
• `"mode": "attribution"` — How the pipeline traces decisions back to behavioral signals
• `"mode": "crosscoder"` — With `"domain"` and `"compare_domain"`, the active features the two domains share and those unique to each; without both, the behavioral patterns that persist across all domains

Step 3: View All Domains

cortex:list_domains({})

Shows all detected project domains with session counts, cognitive style per domain, and last activity timestamps.

Step 4: Rebuild Profiles

When profiles seem stale or after significant new work:

cortex:rebuild_profiles({})

Full rescan of `~/.claude/projects/` session history. Rebuilds all domain profiles, re-extracts patterns, re-classifies cognitive styles, and re-learns behavioral features.

Step 5: Record Session End

Normally automatic via hook, but can be called manually:

cortex:record_session_end({
  "domain": "<current domain>",
  "summary": "<what was accomplished>"
})

Updates profiles incrementally via EMA (Exponential Moving Average) — no full rebuild needed.

Tips

• **Profiles improve over time**: The more sessions you have, the more accurate the cognitive profiling becomes
• **Cross-domain bridges are powerful**: If Cortex detects you use similar patterns in different projects, it surfaces those connections
• **Blind spots are actionable**: If Cortex says you under-use certain tools or miss certain categories, consider addressing those gaps
• **Domain detection is automatic**: `cortex:detect_domain` classifies the current working directory into a known domain — no manual tagging needed