cli-chat — Claude Code skill: CLI chat with OmniRoute
Sends completions to OmniRoute chat models from the terminal, streams responses, and starts an interactive REPL session.
npx skills add cheaper-inference/OmniRoute --skill cli-chat -a claude-codeWhat cli-chat does
The skill adds three commands: omniroute chat [prompt] for one-off requests, omniroute repl for interactive dialogue with session resumption, and omniroute stream [prompt] for streaming output. Supports all OmniRoute providers, combo routing, and system prompt configuration via -s or --system flags. You can pass text from a file (--file), stdin (--stdin), or directly as an argument.
Flags let you specify model (-m), temperature, max-tokens, top-p, reasoning effort (--reasoning-effort), and thinking budget (--thinking-budget). Streaming supports --raw, --debug, and --save to write output to a file. REPL can resume a previous session with --resume <session>, and --no-history disables history in the chat command.
What it helps with
- send a request to an LLM from the terminal without writing a script
- stream model responses in real time
- start an interactive dialogue with history and session resumption
- switch between OmniRoute providers and combo routes
- pass system prompt and generation parameters via flags
Example prompts for the agent
- Send a question from
question.txttogpt-4oand save the answer toanswer.md - Start a REPL with model
claude-3-5-sonnet-20241022and system prompt 'You are a text editor' - Stream a response to 'Explain how async/await works in JS' with temperature 0.7
How to install cli-chat
1Skills CLI (Vercel)
npx skills add cheaper-inference/OmniRoute --skill cli-chat -a claude-code2Manually: put the files in place
~/.claude/skills/cli-chat/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/cli-chat/
1Skills CLI (Vercel)
npx skills add cheaper-inference/OmniRoute --skill cli-chat -a codex2Manually: put the files in place
.agents/skills/cli-chat/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add cheaper-inference/OmniRoute --skill cli-chat -a cursor2Manually: put the files in place
.cursor/skills/cli-chat/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install cheaper-inference/OmniRoute cli-chat2Manually: put the files in place
.github/skills/cli-chat/SKILL.md1Skills CLI (Vercel)
npx skills add cheaper-inference/OmniRoute --skill cli-chat -a gemini-cli2Platform CLI
gemini skills install https://github.com/cheaper-inference/OmniRoute --path skills/cli-chatManually: put the files in place
.devin/skills/cli-chat/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add cheaper-inference/OmniRoute --skill cli-chat -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about cli-chat
How to install cli-chat in Claude Code?
Run npx skills add cheaper-inference/OmniRoute --skill cli-chat -a claude-code, then open ~/.claude/skills/cli-chat/SKILL.md.
What does cli-chat do?
Adds a CLI interface for sending requests to OmniRoute chat models: one-off commands, streaming, and interactive REPL with history.
Which providers does cli-chat support?
All OmniRoute providers, including combo routing via the --combo <name> flag.
Source text
SKILL.md · MIT
--- name: cli-chat description: Send chat completions, stream responses, and start an interactive REPL session from the CLI. Supports all OmniRoute providers, combo routing, and system prompt configuration. --- <!-- generated by src/lib/agentSkills/generator.ts; manual edits will be overwritten --> ## Overview Send chat completions, stream responses, and start an interactive REPL session from the CLI. Supports all OmniRoute providers, combo routing, and system prompt configuration. ## Quick install ```bash npm install -g omniroute # or: npx omniroute omniroute --version ``` ## Subcommands ### `chat [prompt]` **Flags:** - `--file <path>` - `--stdin` - `-s, --system <prompt>` - `-m, --model <id>` - `--max-tokens <n>` - `--temperature <t>` - `--top-p <p>` - `--reasoning-effort <level>` - `--thinking-budget <tokens>` - `--combo <name>` - `--responses-api` - `--stream` - `--no-history` **Example:** ```bash omniroute chat [prompt] ``` ### `repl` **Flags:** - `-m, --model <id>` - `--combo <name>` - `-s, --system <prompt>` - `--resume <session>` **Example:** ```bash omniroute repl ``` ### `stream [prompt]` **Flags:** - `--file <path>` - `--stdin` - `-m, --model <id>` - `-s, --system <prompt>` - `--combo <name>` - `--max-tokens <n>` - `--responses-api` - `--raw` - `--debug` - `--save <path>` **Example:** ```bash omniroute stream [prompt] ```
© cheaper-inference · MIT · cheaper-inference/OmniRoute
