Roleplay a stakeholder from the course's running campus-cafeteria case so a student can practise requirements-elicitation interviewing (asking 'why', 5-Whys, separating symptoms from root causes, spotting contradictions). Calling this turns YOU into the character: stay fully in character, in Hebr...
AI agents call interview_stakeholder to retrieve information from Sad without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
| Parameter | Type | Required | Description |
|---|---|---|---|
persona | string | — | Which stakeholder to become. Defaults to 'shimon' (the cafeteria manager). |
user_question | string | — | The student's original request exactly as they typed it. Always pass this for analytics. |
Parameters from the server's own tool schema.
Even though interview_stakeholder only reads data, uncontrolled read access leaks sensitive information and racks up API costs — an agent caught in a retry loop can make thousands of calls a minute without anyone noticing.
Risk signalsBulk/mass operation — affects multiple targets · Admin/system-level operation
Attacks that exploit this kind of access
Roleplay a stakeholder from the course's running campus-cafeteria case so a student can practise requirements-elicitation interviewing (asking 'why', 5-Whys, separating symptoms from root causes, spotting contradictions). Calling this turns YOU into the character: stay fully in character, in Hebrew first person, and reveal root causes only when the student asks good questions. Currently available: 'shimon' (שמעון, the cafeteria manager who initiated the project). The teacher steers with bracketed commands like [צא מהדמות]/[out of character] or [רמז]. A student working alone ends by typing [סיכום] (or saying they're done) to get a formative, ungraded coaching debrief on their interview — coverage of the root causes, questioning technique, what they missed and how, with an option to go back in ([חזור לדמות]). Call once to summon the character, then answer every following message in character until released. It is categorised as a Read tool in the Sad MCP Server, which means it retrieves data without modifying state.
interview_stakeholder accepts 2 parameters: persona, user_question. The full parameter table on this page comes from the server's own tool schema.
Register the Sad MCP server in PolicyLayer and add a rule for interview_stakeholder: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Sad. Nothing to install.
interview_stakeholder is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the interview_stakeholder rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for interview_stakeholder. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
interview_stakeholder is provided by the Sad MCP server (sad-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →