Slash Commands

Slash commands are typed directly in the chat input during a session. They control session behavior, model selection, and context management.

Command Reference

Command	Description
/model	Switch between configured models at runtime
/mcp	Connect configured MCP servers and list discovered tools
/thinking	Control thinking/reasoning depth per model
/skills	Enable/disable skills with a checkbox picker
/resume	Resume a previous session from its log
/summarize	Summarize older context segments to free up space
/compact	Full context reset with a continuation summary

/model

Opens a picker showing all configured models. Select one to switch the active model for the remainder of the session.

/model

Models come from the providers you set up during longeragent init. If you need a model that is not listed, re-run the init wizard to add more providers. For GLM, Kimi, and MiniMax, selecting a model with a missing key can prompt you to import a detected env var or paste the key directly.

Inline API-key arguments such as /model openai:gpt-5.4 key=... or api_key=... are no longer supported. Use the picker flow or re-run longeragent init.

See Model Switching for details.

/mcp

Connects the MCP servers configured in ~/.longeragent/mcp.json and prints the discovered tool list.

/mcp

This is useful as a quick health check before your first agent turn. If a server is misconfigured or not exposing tools, /mcp will show that immediately.

See MCP Integration for details.

/thinking

Opens a picker showing the available reasoning levels for the current model.

/thinking

Levels vary by provider. For example, Anthropic Claude 4.6 supports: off, low, medium, high, max. OpenAI models support: none, low, medium, high, xhigh.

See Model Switching for the full table.

/skills

Opens a checkbox picker where you can toggle skills on or off.

/skills

Enabled skills are loaded as dynamic tools that the agent can use. Disabled skills are ignored.

See Skills for details.

/resume

Resume a previous session from its session log.

/resume

LongerAgent keeps session logs that record every runtime event. The /resume command lets you pick a previous session and continue from where it left off.

/summarize

Summarize older context segments to free up space. You can optionally provide instructions about what to preserve.

/summarize
/summarize Keep the auth refactor details
/summarize Preserve file references and test results

Summarization compresses selected segments while keeping key decisions and findings intact. It is a lighter alternative to /compact -- use it when context is growing but you do not need a full reset.

See Context Management for details.

/compact

Full context reset with a continuation summary. You can optionally provide instructions about what to preserve.

/compact
/compact Preserve the DB schema decisions

After compaction, the agent starts with a fresh context window containing only the continuation summary, the Important Log, and AGENTS.md files. Use this when the session is getting slow or when you want a clean slate.

See Context Management for details.