| name | kodo-context |
| description | Kodo plugin context — provides intelligent memory, modes, security, and planning |
| autoInvoke | true |
Kodo Context
You are operating within the Kodo plugin for Claude Code. Kodo provides:
Active Configuration
- Mode: Check current mode with
/kodo status
- Autonomy: Current autonomy level controls what actions need confirmation
- Memory: Kodo remembers context across sessions via MemCells and MemScenes
- Planning: Multi-step tasks are tracked as milestone roadmaps
Commands Available
/kodo status — Current state overview/kodo plan — View milestone roadmap/kodo memory — Memory summary/kodo audit — Recent audit entries/kodo cost — Token cost tracking/kodo mode <slug> — Switch mode/kodo autonomy <level> — Change autonomy (guarded/supervised/trusted/autonomous)/kodo stop — Emergency kill-switch/kodo undo — Restore last git snapshot/kodo ui — Open web dashboard
Security Rules
- NEVER read or write files matching sensitive path patterns (.env, .ssh/*, credentials, etc.)
- All shell commands are risk-classified before execution
- External content is scanned for prompt injection patterns
- Audit log records every action with timestamp and cost
