Skip to main content
Glama

privacy_session_start

Start an isolated, throwaway browser session where nothing is persisted to memory, audit logs, or events. Use it to browse or research with local-only storage, optionally with VPN, private browsing, or local content redaction.

Instructions

Start a privacy session (own throwaway browser context - nothing persisted).

mode='standard' isolated context (incognito-equivalent). mode='private' the same isolation plus: the VPN you asked for (call privacy_vpn_connect() - it opens and connects the app, with your approval), Brave (or Chromium) as the engine in a private window, and the AI assistant (Venice AI) opened first as the research tab while the real sites stay in a second tab. Call body_privacy_preflight() first: it reports what is installed, what is running, and what you should switch on.

keep_content_local=True the page TEXT stays on this machine: it is redacted and summarised by a local model and only that comes back. It needs a local model (Ollama + ollama pull qwen3:0.6b); without one the text is withheld rather than sent to your model provider.

Nothing from this session is written to persistent memory, the audit log or the event stream. It reduces local persistence only - it is NOT anonymity: websites, network operators, device administrators and downloaded files can still observe it.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "privacy_session_start"
  }
}

Input Schema

NameRequiredDescriptionDefault
vpnNo
modeNostandard
labelNosession
browserNochromium
show_ipNo
assistantNo
keep_content_localNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.4.0

TDQS

A3.8/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations supplied, the description carries the full disclosure burden and does so well: it states what is not written (persistent memory, audit log, event stream), the local-model dependency and its failure mode ('text is withheld rather than sent to your model provider'), and an explicit non-anonymity caveat about websites and network operators observing the session. It does not explain what the call returns (session handle or ID) or the session's lifespan.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Purpose is front-loaded in one sentence, and the rest is organized as parameter-keyed blocks, which makes scanning efficient. It runs long and the indented multi-line mode='private' block is denser than necessary, but essentially every sentence carries a distinct behavioral fact.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 7-parameter, zero-annotation tool with no output schema, the description covers isolation semantics, persistence, prerequisites, and a limitation disclaimer well. Missing pieces are the undocumented params (label, show_ip, browser, assistant as standalone options) and any statement about the returned session handle or how this differs from privacy_session_open.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 0%, so the description must supply all parameter meaning, and it only partially does. It gives real semantics for mode (standard vs private), vpn, keep_content_local, and implicitly browser and assistant, but label and show_ip are never mentioned, and mode's two values are not declared as the complete set. Substantive value is added for the most consequential parameters, but over half the surface remains undocumented.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The opening sentence names a specific verb and resource ('Start a privacy session') and immediately bounds the scope ('own throwaway browser context - nothing persisted'). The mode breakdown makes the two operating flavors concrete. It stops short of distinguishing itself from the nearby privacy_session_open sibling, which an agent would have to disambiguate on its own.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Includes an explicit prerequisite: 'Call body_privacy_preflight() first', and names privacy_vpn_connect() as the VPN mechanism with the approval requirement noted. The conditional guidance for keep_content_local (needs Ollama, otherwise text is withheld) tells the agent when the option is viable. No explicit when-not-to-use or comparison to sibling session tools.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.