CLIver MCP server
Commit Nov 11, 2026An embeddable Python agent core that provides an agent loop, tool calling, skills, and media generation capabilities.
Also known as cliver-project/CLIver
- Stars
- 8
- Tools
- —
- Downloads, 30 days
- —
- Latest version
- —
- Last commit
- Nov 11, 2026
- License
- Apache-2.0
What is CLIver MCP?
CLIver is an embeddable Python agent core that provides an agent loop, tool calling, skills, and media generation capabilities.
Summary generated from the project's README
About CLIver MCP server
Integrating an autonomous agent loop into Python applications is what CLIver enables through its embeddable core library. Developers can use the cliver-core package to implement tool calling, model bindings for OpenAI and Anthropic protocols, and various media generation capabilities including text-to-speech and image generation. The library operates as a set of Python objects that take messages as input and return responses, requiring API keys for chosen providers like DeepSeek.
Users can also interact with the system via a companion CLI application for running one-shot tasks or using a streaming interactive REPL. This command-line interface supports tool approval modes and skill discovery. The system includes nine built-in tools such as shell access, code execution, and web fetching, alongside support for Model Context Protocol clients via stdio or HTTP. It manages complex workflows through execution policies, permission gates, and structured output validation.
What you can do with it
- 1Automate local file operationsA developer can instruct the agent to read a configuration file and summarize its contents using built-in file I/O tools.
- 2Build custom agentic workflowsSoftware engineers can embed the agent loop directly into their own Python code to handle chat and media tasks.
- 3Manage multi-modal AI tasksAn AI application can utilize the core to handle diverse media requirements like speech-to-text or video generation.
- 4Execute terminal commands safelyUsers can run shell commands or code execution through the CLI while providing interactive approval for sensitive actions.
Written from the project's README. Every use case is checked against a quote from it.
Configuration
CLIVER_MODELCLIVER_PROVIDERDEEPSEEK_API_KEY
Automated checks
Ran Nov 14, 2026Full report →Not verified yet. A server needs a score of 80, no serious finding, and enough data for us to check. How checks work · What to fix for more points
- Source repository linked and reachablecliver-project/cliver is public and reachablegithub.com/cliver-project/cliver ↗Pass
- Package names this repository as its sourceNot checked
- Build provenance (Sigstore attestation)Not checked
- Open source license (OSI)Apache-2.0spdx.org/licenses/Apache-2.0.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-11-11github.com/cliver-project/cliver ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNot checked
- No critical advisory in the package or its dependenciesNot checked
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubCLIver is an embeddable agent core with a companion CLI, built from two packages:
cliver-core— the embeddable engine: an agent loop with tool calling, provider engines for the OpenAI and Anthropic protocols, per-purpose model bindings (chat, image, video, TTS, ASR, embed, rerank), nine built-in tools, Agent Skills, MCP, and opt-in permission and execution-policy axes. No CLI, database, or config files — Python objects in, messages out.cliver— the CLI application built oncliver-core: config-path and skill-discovery conventions plus aclivercommand that runs one task per invocation or starts a streaming interactive REPL, with interactive tool approval in both modes. It remains experimental and is not on PyPI yet.
cliver-core 0.3 is the first stable series: the public API stays backward compatible, and
0.3.x patch releases contain bug fixes and backward-compatible additions. The cliver CLI is
for tinkering:
run one-shot tasks or a multi-turn REPL with your own skills and tools, approve mutating calls
interactively, and inspect tool events. See
Stability & Versioning
for the support policy.
Quick Start
Install the core library (Python 3.10–3.13):
pip install cliver-core==0.3.1
As a library
import asyncio
from cliver_core import AgentCore, ModelBindings, presets
async def main():
bindings = ModelBindings(llm=presets.deepseek(api_key="..."))
async with AgentCore.open(bindings) as agent:
response = await agent.chat("Summarize the agent loop in one sentence.")
print(response.message.content)
asyncio.run(main())
See the cliver-core README for tools, skills, MCP, media capabilities, and the full API.
From the terminal
The cliver CLI lives in this repository (not on PyPI yet). With a prompt it runs one task;
with no prompt on a terminal it starts a streaming REPL:
export CLIVER_MODEL=deepseek-v4-flash
export CLIVER_PROVIDER=deepseek # or CLIVER_BASE_URL for any compatible endpoint
export DEEPSEEK_API_KEY=... # or CLIVER_API_KEY
uv run cliver --yes "Read pyproject.toml and summarize the dependencies."
uv run cliver --yes # interactive REPL: /help, /clear, /exit
--yes approves ASK tools without prompting; without it, safe tools still run and each ASK
tool asks on the terminal (non-interactive runs deny them). --skill NAME runs a skill
explicitly, --cwd DIR selects the project root for skill discovery, and {config_dir}/.env
is loaded automatically. In the REPL, each turn streams its answer and keeps the conversation
for follow-ups.
Built-in Capabilities
- Model-agnostic — OpenAI and Anthropic protocols, live-verified presets, any compatible endpoint
- Built-in tools — file I/O, shell, code execution, grep, directory listing, todo tracking, and web fetch
- Skills system — Agent Skills loading plus model-driven
FindSkills/ActivateSkill/ReadSkillFile - MCP support — stdio and streamable HTTP clients with CLIver-style discovery
- Permission and execution policy — opt-in SAFE/ASK gates and env/cwd/path scopes
- Media capabilities — image, video, text-to-speech, speech-to-text, embedding, and rerank
- Reliability — same-purpose failover, run deadlines, and explicit credential rotation
- Structured output — schema-validated
chat_structured()with retries and usage accounting - Context management — declarative compaction plans, an opt-in
LlmCompactor, and a full-history escape hatch, all inside the run deadline - System One — decision-model tool/skill selection with a fail-open selector
- Embeddable API —
AgentCore.chat(),.stream(), and.run_skill()with no CLI dependencies
Repository Layout
packages/cliver-core/ # embeddable library (published)
packages/cliver/ # CLI application built on cliver-core
website/ # documentation site (Astro + Starlight)
Documentation
Full documentation at cliver-project.github.io/CLIver
Development
git clone https://github.com/cliver-project/CLIver.git
cd CLIver
make init # Set up dev environment
make test # Run the workspace test suite
make lint # Lint and format check
make format # Auto-fix lint/format issues
make dist # Build the cliver-core sdist + wheel
cd website && npm install && npm run dev # Docs dev server at localhost:4321
License
CLIver MCP questions
Does CLIver support MCP?
Yes, it includes MCP support with stdio and streamable HTTP clients featuring CLIver-style discovery.
Can I use CLIver with different AI providers?
Yes, it is model-agnostic and supports OpenAI and Anthropic protocols, as well as any compatible endpoint.
What built-in tools are available in CLIver?
The library includes file I/O, shell, code execution, grep, directory listing, todo tracking, and web fetch.