Mastra and Deedglass
Give a Mastra agent Deedglass’s tools in one file. The file starts Deedglass’s MCP server (npx -y deedglass mcp serve) and hands its tools to the agent: each call goes through your policy, the risky ones wait for a person, and each one is sealed as a deed you can check later.
Tested with @mastra/core 1.79.0 and @mastra/mcp 2.5.0 (2026-10-21).
Before you start
Section titled “Before you start”- In any folder, run
npx deedglass init(Docker, about 15 seconds; the quickstart). It starts Deedglass and a demo Postgres with 500 orders, and saves an agent key in.deedglass/quickstart.key. - Run the file below from that folder (or a folder inside it).
deedglass mcp servelooks for the key there; from anywhere else, setDEEDGLASS_API_KEYandDEEDGLASS_SERVERyourself (how mcp serve finds them). - Under the
postgres-safepolicy thatinitturns on, reads run,DROPandTRUNCATEare refused, and a write waits for a person in Console › Approvals before it runs.
The MCP route
Section titled “The MCP route”// Run where you ran `npx deedglass init`. Needs OPENAI_API_KEY.import { Agent } from '@mastra/core/agent';import { MCPClient } from '@mastra/mcp';
const mcp = new MCPClient({ id: 'deedglass', servers: { deedglass: { command: 'npx', args: ['-y', 'deedglass', 'mcp', 'serve'] } } });try { const tools = await mcp.listTools(); // deedglass_postgres_query and so on const agent = new Agent({ id: 'orders', name: 'Orders', instructions: 'Answer questions about orders.', model: 'openai/gpt-5.6-sol', tools }); console.log((await agent.generate('How many rows are in public.orders?')).text);} finally { await mcp.disconnect();}Install and run:
npm install @mastra/core@1.79.0 @mastra/mcp@2.5.0 zod@4.6.7node agent.tsModel key: OPENAI_API_KEY (the model string is the one in Mastra’s agent docs). Any model in Mastra’s router format works.
Tool names: Mastra prefixes each tool with its server’s key, so the tools arrive as deedglass_postgres_query and so on. Deedglass lists its tools with names every model API takes: postgres_query, postgres_execute, deedglass_check_approval (letters, digits, _ and -; Anthropic’s rule is ^[a-zA-Z0-9_-]{1,128}$ and OpenAI’s allows the same characters up to 64). Your policies and the sealed records keep the dotted name, postgres.query; Deedglass maps one to the other (MCP.md). Deedglass 0.7 and earlier listed the dotted names, which these APIs refuse.
What you see: the answer (500 rows), and in Console › Ledger or npx deedglass actions list one sealed postgres.query action per call, with the rule that allowed it. If the agent tries a write, the tool answers that it is waiting for approval and nothing has run; once someone approves, deedglass_check_approval lets the agent carry on. A refused call answers with the policy’s sentence and a record id.
First start: npx -y deedglass can take several seconds the first time. If a framework’s MCP connect timeout is shorter, run npx -y deedglass version once first, or install the CLI (INSTALL.md) and use deedglass as the command.
Check it without a model key
Section titled “Check it without a model key”examples/frameworks/mastra/ has the file, its pinned packages and a check file that needs no model
key: it starts deedglass mcp serve through the same framework client, lists its tools and builds the Agent with them as the file does. From the folder where
init ran:
path/to/deedglass/examples/frameworks/run.sh mastra # install the pins in a temp folder, run the checkpath/to/deedglass/examples/frameworks/run.sh mastra --agent # also run the agent file when its model key is setThe workflow frameworks.yml runs every check against a fresh init stack each week and on pull
requests that touch these files. It isn’t a required check: a framework release that doesn’t install
is reported as skipped, not as a failure.
Sources and what isn’t verified
Section titled “Sources and what isn’t verified”Every framework API in the files above comes from the framework’s own docs, checked 2026-10-21:
new MCPClient({ id, servers: { name: { command, args } } }),await mcpClient.listTools(),disconnect(): https://mastra.ai/docs/connections/mcpnew Agent({ id, name, instructions, model, tools }),agent.generate(…),.text: https://mastra.ai/docs/agents/overviewrequireToolApprovalper MCP server, to ask your own user before Deedglass sees the call: https://mastra.ai/docs/agents/human-in-the-loop- Tool-name rules: the MCP spec’s tool names (https://modelcontextprotocol.io/specification/2025-11-25/server/tools#tool-names) and Anthropic’s tool definitions (https://platform.claude.com/docs/en/agents-and-tools/tool-use/define-tools). OpenAI’s API reference returned 403 when we checked; its rule here is the one
@deedglass/sdk’s OpenAI adapter already follows.
Not verified:
- Mastra’s docs register agents in a
Mastrainstance and callgetAgentById; this file callsgenerate()on the agent directly. - That Mastra calls a tool by its key in the tools object with a live model (no model key in CI).