From cortex
Establishes that Cortex context must be consulted before answering code questions, and dispatches to the right Cortex skill for research, impact, and review tasks.
How this skill is triggered — by the user, by Claude, or both
Slash command
/cortex:using-cortexThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Cortex maintains a local, ranked index of this repository (graph, embeddings,
Cortex maintains a local, ranked index of this repository (graph, embeddings, rules). Never answer a code question from memory when Cortex is available.
Before answering any question about this codebase, run:
cortex search "<query>" --json
Cite the entity ids and paths you used. If a command fails because there is
no .context/ directory or no index, say so and suggest cortex bootstrap
(first time) or cortex update (stale index) instead of guessing.
| Situation | Skill / command |
|---|---|
| "How does X work?", exploring code | repo-research skill |
| Planning a refactor or risky change | change-impact skill |
| Reviewing a file or diff | pattern-review skill |
| Finalizing a PR or full review | context-review skill |
| Architectural constraints | cortex rules --json |
| After significant changes | cortex update |
If Cortex MCP tools are connected, context.search, context.get_related,
context.get_rules, context.impact, and context.reload map to the CLI
commands above. Prefer whichever is available; the CLI needs no registration.
npx claudepluginhub danielblomma/cortex --plugin cortexExplores repository structure and code by retrieving Cortex context, locating implementations, and answering 'how/where does X' questions.
Explores codebases indexed by Repowise using MCP tools for overview, Q&A, symbol search, and context retrieval. Activates for architecture questions and structural understanding.
Navigates codebases indexed by rag-rat using MCP tools for semantic search, symbol lookup, impact analysis, and repo memories.