Context MCP Server

Recursive Claude spawning for autonomous project builds with self-healing test loops

Local serverstdio

What is the Context MCP server?

Context MCP server exists for a simple reason — assistants are far more useful when they can act on Context directly instead of describing what you should do. Recursive Claude spawning for autonomous project builds with self-healing test loops.

What you get

Autonomous build loop that plans, builds, reviews, and learns.

Setting it up

Setup follows the usual MCP pattern — install or clone the server, register it in your client's configuration file, restart the client.

What the assistant can call

Once Context is connected, these are the calls the assistant has available:

  • Position — Letter
  • suffix!
  • Mode — Behavior
  • Sprint — Runs all tasks, then stops
  • Review — Runs one task at a time, creates a PR per task, pauses for approval
  • Coach — Runs an intake pre-flight that clarifies SPEC.md before Scout, then proceeds like Auto
  • Service — Unattended build-service mode: runs all tasks then stops; a WIP/audit-failed task is terminal (no retry)
  • Both — Two complete independent pipelines run in parallel, one per provider
  • Field — Default
  • sandbox — true
  • sandbox_image — "foundry-sandbox:latest"
  • sandbox_extra_mounts — []

Configuration and credentials

You will need one environment variable: ANTHROPIC_API_KEY. The server will not start without them, which is usually why the tools fail to appear on a first run. Keep credentials in your client's env block or a secrets manager rather than in a file you might commit.

Before you rely on it

  • It runs with your machine's permissions. That is convenient and also the reason to think about what you point it at before you approve a tool call.
  • With 14 tools registered it takes up a noticeable share of the context window; turn it off in projects that never touch Context.
  • Missing credentials fail quietly in some clients — if no tools show up, check the environment block first.
  • MCP clients confirm each tool call by default. Leave that on until you have watched what the context mcp server does with a few real requests.

Choosing this one

This sits in the planning and project tracking group, where several servers overlap in what they claim to do but differ sharply once you actually set them up. Context's toolset — Position, suffix, Mode and 11 more — is a fair guide to whether it matches your workflow. It is maintained by context-foundry; worth a glance at recent repository activity before you build anything load-bearing on it.

This entry was verified against Context's own documentation before publication; SyncDev keeps the directory reviewed rather than auto-generated.

Available tools

ToolWhat it does
PositionLetter
suffix**!**
ModeBehavior
SprintRuns all tasks, then stops
ReviewRuns one task at a time, creates a PR per task, pauses for approval
CoachRuns an intake pre-flight that clarifies SPEC.md before Scout, then proceeds like Auto
ServiceUnattended build-service mode: runs all tasks then stops; a WIP/audit-failed task is terminal (no retry)
BothTwo complete independent pipelines run in parallel, one per provider
FieldDefault
sandboxtrue
sandbox_image"foundry-sandbox:latest"
sandbox_extra_mounts[]
MetricWhere
InjectedPatterns panel + stats row

Configuration

VariableDescriptionRequired
ANTHROPIC_API_KEYCredential the server authenticates with.Yes

Example prompts to try

  • Use Context to Position.
  • Use Context to suffix.
  • Use Context to Mode.

Frequently asked questions

It connects Context to MCP-compatible AI assistants such as Claude and Cursor, exposing 14 tools (Position, suffix, Mode, and more) that the assistant can call on your behalf. Instead of copying data back and forth by hand, the assistant works with Context directly.