PromptSpeak Governance
Block or validate agent tool calls against governance rules before they execute in your coding workflow.
Overview
PromptSpeak is a Ship-phase MCP server that provides pre-execution governance by validating AI agent tool calls before they execute.
What is this MCP server?
- PromptSpeak Governance MCP server v0.4.1 (@chrbailey/promptspeak-mcp-server) with stdio npm transport.
- Pre-execution governance: validates tool calls before the agent runs them.
- Policy layer for Claude Code–style agents that can delete files, call APIs, or run shell commands.
- npm-installable MCP server from chrbailey/promptspeak-mcp-server repository.
- Complements post-hoc review by stopping disallowed actions at invocation time.
- MCP server version 0.4.1 published as @chrbailey/promptspeak-mcp-server on npm
- stdio transport; manifest lists npm package without required secret env vars
- Described explicitly as pre-execution governance validating tool calls before execute
What problem does it solve?
Prompt-only safety fails once agents can run real tools, and solo builders cannot manually supervise every invocation.
Who is it for?
Indie developers running powerful agent toolkits who need enforceable pre-flight checks on MCP and native tools.
Skip if: Static code scanners, dependency CVE dashboards, or teams with no MCP tool execution in their workflow.
What do I get? / Deliverables
After you register PromptSpeak, tool calls pass through governance validation first so disallowed operations are caught before execution.
- Registered PromptSpeak Governance MCP server v0.4.1 on stdio
- Pre-execution validation step on agent tool calls
- Reduced risk of out-of-policy automated actions before they hit systems
Recommended MCP Servers
Journey fit
How it compares
Pre-execution policy MCP, not a post-merge code review skill or vulnerability database.
Common Questions / FAQ
Who is PromptSpeak for?
Solo builders and small teams using agent tool execution who want server-side governance instead of relying only on system prompts.
When should I use PromptSpeak?
Use it during Ship security setup—and ongoing Operate agent work—whenever tools can mutate repos, infra, or external services.
How do I add PromptSpeak to my agent?
Install @chrbailey/promptspeak-mcp-server from npm, register it as a stdio MCP server in Claude Code or Cursor, and configure your governance policies per PromptSpeak docs.