Tool Policies
Field-level control over what an agent can read and change through StackOne: blocks tools, masks fields, redacts PII and runs plain-language semantic rules that a decision model answers on every call, on inputs and on responses.
Find alternatives to Tool Policies
What it does
- Guardrails and safety
Sources
- Pricing
-
Price evidence read 2026-11-16: Tool Policies is listed under the Team plan card ("Everything in Starter, plus:"), which is a paid plan, with Enterprise above it by contact; Starter, the free plan, does not list it, so no free model is recorded (standard 4.1 row 2).
- Description
-
Official documentation · 16 Nov 2026
Page text fetched 2026-11-16 (curl, full text read); page quote: "Write the rule as a plain-language question. A decision model answers it for each field or payload, on inputs before the tool runs and on responses before the agent sees them" "Control what agents can read and change, down to individual fields and input values.".
- What it does
-
Official documentation · 16 Nov 2026
Page text fetched 2026-11-16 (curl, full text read); page quote: "Write the rule as a plain-language question. A decision model answers it for each field or payload, on inputs before the tool runs and on responses before the agent sees them" "Control what agents can read and change, down to individual fields and input values.".
- Sold within
-
The pricing page lists Tool Policies under the Team plan card, which adds it to Starter: "Control what agents can read and change, down to individual fields and input values." Quote of the card heading: "Everything in Starter, plus:".
- Name
-
Official documentation · 16 Nov 2026
Page text fetched 2026-11-16 (curl, full text read); page quote: "Write the rule as a plain-language question. A decision model answers it for each field or payload, on inputs before the tool runs and on responses before the agent sees them" "Control what agents can read and change, down to individual fields and input values.".
- URL
-
Official documentation · 16 Nov 2026
Page text fetched 2026-11-16 (curl, full text read); page quote: "Write the rule as a plain-language question. A decision model answers it for each field or payload, on inputs before the tool runs and on responses before the agent sees them" "Control what agents can read and change, down to individual fields and input values.".
Also from StackOne
-
Tool Gateway MCP Developer tool
Tool Gateway MCP is one MCP endpoint that gives AI assistants governed access to 630+ business apps, each under the signed-in person's own permissions.
-
AI Connector Builder Developer tool
Builds or patches a connector to any API from a prompt or an OpenAPI spec and returns tested actions that agents call over MCP, the AI SDKs or the Actions API.
-
StackOne Defender Developer tool
Open-source library that detects and blocks indirect prompt injection hidden in documents, emails and tickets before tool responses reach the model, also built into every StackOne connector.
-
Tool Discovery Developer tool
Semantic tool search for agents: finds the right action among tens of thousands by meaning with a fine-tuned retrieval model, so only the schemas the agent needs are loaded into its context.
-
PII Redaction Application
Detects personal data such as email addresses, phone numbers, SSNs and card numbers in an agent's tool inputs and responses with StackOne's own PII model, then redacts the span or blocks the call, without relying on field names.
Open StackOne in the directory
Something wrong here? Send a correction — quote this product id: stackone-tool-policies.