version: "1.0.0" name: dev-mcp description: MCP development for Capsem. Covers the capsem-mcp host MCP server (AI agent sandbox control via stdio), the guest MCP relay and host MITM MCP endpoint (tool routing to external servers via framed vsock), and using capsem MCP tools for fast debugging. Use when working on the MCP server, endpoint, tool routing, policy evaluation, mcp_calls telemetry, or when you need to debug anything inside a VM. Also use this skill when capsem MCP tools are available and you want to understand the fastest way to test changes interactively.

MCP in Capsem

Capsem has two MCP components:

1. capsem-mcp (host): MCP server over stdio that lets AI agents (Claude Code, Gemini CLI) control sandboxes -- create/delete VMs, exec commands, read/write files, query telemetry. Bridges to capsem-service HTTP API over UDS.
2. Guest MCP relay + MITM MCP endpoint: bridges AI agents running inside a guest VM to external MCP servers on the host via framed MCP records over vsock port 5002.

Using capsem MCP tools for fast debugging

When the capsem MCP server is configured in your AI CLI, you have direct VM control without leaving the conversation. This is the fastest debug loop for any in-VM work.

Available tools

Debug workflow

-- Quick one-shot (no VM management needed):
capsem_run { command: "capsem-doctor -k net" }
 
-- Iterative debugging (long-lived VM):
1. capsem_create        -- boot a fresh sandbox (add name for persistence)
2. capsem_exec          -- run the thing you want to test
3. capsem_read_file     -- check config, logs, state
4. capsem_inspect       -- query telemetry tables
5. (fix code on host, rebuild with `just build`)
6. capsem_delete        -- tear down
7. repeat from 1

Common debug patterns

Verify a guest command works:

capsem_exec { id: "vm-1", command: "capsem-doctor -k net" }

Check network policy enforcement:

capsem_exec { id: "vm-1", command: "curl -s https://blocked-domain.com" }
capsem_inspect { id: "vm-1", sql: "SELECT domain, action, status_code FROM net_events ORDER BY timestamp DESC LIMIT 10" }

Verify telemetry pipeline:

capsem_inspect { id: "vm-1", sql: "SELECT server_name, tool_name, decision, duration_ms FROM mcp_calls ORDER BY timestamp DESC" }
capsem_inspect { id: "vm-1", sql: "SELECT COUNT(*) as n, operation FROM fs_events GROUP BY operation" }

Read guest config/state:

capsem_read_file { id: "vm-1", path: "/etc/resolv.conf" }
capsem_read_file { id: "vm-1", path: "/tmp/capsem-init.log" }

Write a test script and run it:

capsem_write_file { id: "vm-1", path: "/tmp/test.sh", content: "#!/bin/bash\necho hello" }
capsem_exec { id: "vm-1", command: "chmod +x /tmp/test.sh && /tmp/test.sh" }

When to use MCP tools vs just recipes

MCP tools are for fast, targeted checks during development. Just recipes are for comprehensive validation before committing.

capsem-mcp (host MCP server)

Architecture

AI Agent (Claude Code) <-> capsem-mcp (stdio, rmcp) <-> HTTP/UDS <-> capsem-service

Uses the rmcp crate with #[tool_router] macro for tool definitions. Stateless -- creates a fresh HTTP connection to ~/.capsem/run/service.sock per request.

Parameter conventions

MCP tools use camelCase on the wire (ramMb, cpuCount) because that is the MCP/JSON convention. The capsem-service HTTP API uses snake_case (ram_mb, cpus). The conversion happens inside each tool method -- the #[serde(rename)] attributes on param structs handle deserialization, and the tool builds a new JSON body with the service's field names.

Key source files

Configuration

Registered in AI CLI settings:

{ "mcpServers": { "capsem": { "command": "target/debug/capsem-mcp" } } }

Environment variables

MCP subprocess architecture

The guest MCP path is not a single process. capsem-process (the per-VM host process) owns the MITM MCP endpoint and spawns two privilege-isolated subprocesses that together handle MCP traffic from the guest:

Rationale: isolating external-server connections in a low-privilege subprocess means a compromised third-party MCP server cannot reach the host filesystem or the session DB. The built-in tool server runs in its own process for the same reason.

Wire protocol between capsem-process and the aggregator: length-prefixed msgpack frames on stdio ([4-byte big-endian length][msgpack payload]). Between the aggregator and the built-in server: stdio MCP (standard JSON-RPC per line). Between the in-guest AI agent and capsem-process: /run/capsem-mcp-server relays stdio JSON-RPC as bounded framed MCP records over vsock port 5002. MCP calls pass through the MITM parser/interpreter and write MITM-owned mcp_calls.

Binaries land in ~/.capsem/bin/ at install time: capsem-mcp-aggregator, capsem-mcp-builtin.

Guest MCP Endpoint

The guest MCP relay bridges AI agents in the guest VM to the host MITM MCP endpoint. It runs over vsock port 5002 using bounded length-prefixed MCP frames that carry JSON-RPC payloads and per-frame process attribution.

Framed guest MCP over vsock:5002 must be tested as the default transport, not as an opt-in benchmark mode. The minimum hardening matrix for that path is:

• parser/interpreter: bounded frames, invalid JSON, malformed flags, stream-id reuse, notification/request-id mismatch
• dispatch: initialize, tools/list, builtin tools/call, configured external stdio tools/call, resources/list, prompts/list, and method error mapping
• policy: live policy mutation, per-tool block, resource URI rule, argument-name rule, argument-value rule, return-value rule, deny-over-allow precedence, and proof that blocked requests/responses do not leak original data
• telemetry: session.db rows for success, denial, timeout, process attribution, request/response previews, policy fields, and terminal errors
• boundary: aggregator remains DB-free; MITM/process owns MCP audit writes
• VM E2E: boot a real VM, run /run/capsem-mcp-server with no transport override, then query session.db

Architecture

Guest (Claude/Gemini) -> capsem-mcp-server (stdin/stdout relay)
  -> framed vsock:5002 -> MITM MCP endpoint (capsem-core)
  -> Policy check -> Route to: builtin tools | external MCP servers (via rmcp)
  -> Telemetry -> session.db mcp_calls table

Wire format

Length-prefixed MCP frames over vsock. Each frame contains a bounded JSON-RPC payload plus a stream id, flags, and sanitized process name.

Handshake

Guest sends NUL-prefixed metadata line first:

\0CAPSEM_META:claude\n

Then JSON-RPC messages:

{"jsonrpc":"2.0","id":1,"method":"initialize","params":{}}
{"jsonrpc":"2.0","id":2,"method":"tools/list"}
{"jsonrpc":"2.0","id":3,"method":"tools/call","params":{"name":"github__search","arguments":{"q":"rust"}}}

Tool namespacing

Tools are namespaced with __ (double underscore) to prevent collisions:

• github + search_repos -> github__search_repos
• builtin + http_get -> builtin__http_get

The endpoint parses the namespace to route to the correct server.

Supported methods

Policy evaluation

1. Blocked servers list (highest priority)
2. Allowed servers whitelist (if non-empty)
3. Per-tool decision map
4. Default fallback (Allow/Warn/Block)

Config hierarchy: corp.toml > user.toml > auto-detected from AI CLI settings.

Decisions: Allow, Warn (log + continue), Block (error -32600).

Built-in tools

Snapshot tools (VirtioFS mode only)

snapshots_list, snapshots_changes, snapshots_create, snapshots_delete, snapshots_revert, snapshots_history, snapshots_compact

HTTP tools (always available)

http_get, http_post, http_put, http_patch, http_delete, http_head

All use namespace prefix builtin (e.g., builtin__http_get).

Endpoint key source files

Telemetry (mcp_calls table)

Every request/response logged with: timestamp, server_name, method, tool_name, request/response preview (256KB cap), decision, duration_ms, error_message, process_name, bytes sent/received.

Read references/mcp-wire.md for the full wire format details.

Testing

Unit tests

cargo test -p capsem-mcp -- param serde roundtrips, UDS path resolution, tool router registration, schema constants.

cargo test -p capsem-core mcp -- gateway, policy, server manager, type serialization.

Integration tests (Python)

The MCP integration tests (tests/capsem-mcp/) are black-box tests that boot a real service + VM and exercise the full MCP protocol over stdio.

Run with: just test-mcp (or pytest tests/capsem-mcp/ -m mcp -v)

Test files:

Fixture architecture:

• capsem_service (session scope) -- spawns capsem-service on isolated temp socket, codesigns binaries on macOS
• mcp_session (per-test) -- fresh capsem-mcp subprocess with JSON-RPC handshake, returns McpSession helper
• shared_vm (session scope) -- one long-lived VM for non-destructive tests, avoids repeated boot overhead
• fresh_vm (per-test factory) -- creates uniquely named VMs with auto-cleanup for destructive tests

McpSession helper (tests/capsem-mcp/conftest.py): wraps capsem-mcp subprocess with JSON-RPC 2.0 protocol. Key methods:

• request(method, params) -- send NDJSON, read response
• call_tool(name, args) -- call tool, assert success, parse JSON content
• call_tool_raw(name, args) -- raw response (no assertions)

In-VM diagnostics

just run "capsem-doctor -k mcp" -- tests tool routing and domain blocking inside the guest.

Manual validation

Boot interactively, run a workload, then inspect telemetry:

just run
# (in another terminal)
just inspect-session <vm_id> "SELECT * FROM mcp_calls"

Or use MCP tools directly (see "Fast debugging" section above) for the same workflow without leaving Claude Code.

Lessons learned

1. Never prepend headers to JSON output. MCP tool responses with format=json must return raw, parseable JSON. Do not wrap JSON in pagination headers, content-length prefixes, or any other text. The snapshots_changes tool broke because paginated_response() prepended "Content length: ...\nShowing: ...\n" to the JSON array, making json.loads() fail. Rule: if a tool offers both text and JSON formats, branch early and return JSON directly without passing through text-oriented helpers like paginated_response().