Call another mcp client from your mcp client. Delegate tasks, offload context windows. An agent for your agent
MCP Inception MCP server exists for a simple reason — assistants are far more useful when they can act on MCP Inception directly instead of describing what you should do. Call another mcp client from your mcp client. Delegate tasks, offload context windows. An agent for your agent.
Ok this is a difficult one. Will take some setting up unfortunately. However, if you manage to make this more straightforward, please send me PR's.
Once MCP Inception is connected, these are the calls the assistant has available:
execute_mcp_client — Ask a question to a separate LLM, ignore all the intermediate steps it takes when querying it's tools, and return the outputexecute_parallel_mcp_client — Takes a list of inputs and a main prompt, and executes the prompt in parallel for each string in the inputnote — wait for this before using this featureexecute_map_reduce_mcp_client — Process multiple items in parallel and then sequentially reduce the results to a single outputTools — The Tools tool exposed by this serverDebugging — Since MCP servers communicate over stdio, debugging can be challenging. We recommend using the [MCPYou will need 2 environment variables: MCP_INCEPTION_EXECUTABLE, MCP_INCEPTION_WORKING_DIR. Keep credentials in your client's env block or a secrets manager rather than in a file you might commit.
Setup follows the usual MCP pattern — install or clone the server, register it in your client's configuration file, restart the client.
Plenty of developer tooling servers cover similar ground. The differences that matter in practice are scope of access and how much setup stands between you and a working tool call. MCP Inception's toolset — execute_mcp_client, execute_parallel_mcp_client, note and 3 more — is a fair guide to whether it matches your workflow. It is maintained by tanevanwifferen; worth a glance at recent repository activity before you build anything load-bearing on it.
We check each listing at SyncDev against the project's documentation before it goes live — if something here drifts out of date, it is a bug worth reporting.
| Tool | What it does |
|---|---|
| execute_mcp_client | Ask a question to a separate LLM, ignore all the intermediate steps it takes when querying it's tools, and return the output. |
| execute_parallel_mcp_client | Takes a list of inputs and a main prompt, and executes the prompt in parallel for each string in the input. |
| note | wait for [this](https://github.com/adhikasp/mcp-client-cli/pull/11) before using this feature |
| execute_map_reduce_mcp_client | Process multiple items in parallel and then sequentially reduce the results to a single output. |
| Tools | The Tools tool exposed by this server. |
| Debugging | Since MCP servers communicate over stdio, debugging can be challenging. We recommend using the [MCP Inspector](https://github.com/modelcontextprotocol/inspector), which is available as a package script: |
| Variable | Description | Required |
|---|---|---|
| MCP_INCEPTION_EXECUTABLE | Configuration value read at startup. | Optional |
| MCP_INCEPTION_WORKING_DIR | Filesystem location the server is allowed to use. | Optional |
Kill hallucinated APIs — version-accurate, up-to-date library documentation injected straight into context.
Microsoft's official browser automation server — drive a real browser through the accessibility tree, no screenshots needed.
GitHub's official server — repos, issues, pull requests, Actions and code security, straight from your assistant.
Issue tracking at the speed of conversation — Linear's official hosted server with OAuth and zero install.
Local repository surgery — status, diffs, commits, branches and history for any repo on disk.
Timezone sanity for AI — current time anywhere and correct conversions, without the model doing date math.