AI agents use update-thinking-model to create or update resources in Tianji — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Tianji environment.
This tool modifies existing thinking models but does not delete them irreversibly (which would be Destructive) or execute arbitrary code (Execute). The ability to update model content could affect downstream reasoning or analysis if an AI agent corrupts or replaces legitimate models with malicious ones, warranting medium severity. Confidence is slightly reduced (0.85 vs.
From the tool's definition Tool name 'update-thinking-model' combined with description '更新现有思维模型的内容' (update existing thinking model content) indicates modification of stored data. The verb 'update' is characteristic of Write operations that create or modify data reversibly.
Documented attack patterns abuse exactly the kind of access update-thinking-model gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Tianji, and nothing reaches the server without passing your rules. This is the rule we recommend for update-thinking-model:
{
"version": "1",
"default": "deny",
"tools": {
"update-thinking-model": {
"limits": [
{
"counter": "update-thinking-model_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}update-thinking-model stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
更新现有思维模型的内容. It is categorised as a Write tool in the Tianji MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Tianji MCP server in PolicyLayer and add a rule for update-thinking-model: 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 Tianji. Nothing to install.
update-thinking-model is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the update-thinking-model 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 update-thinking-model. 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.
update-thinking-model is provided by the Tianji MCP server (lanyijianke/thinking_models_mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Tianji, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
19 Tianji tools catalogued and risk-classified — across an index of 43,000+ MCP servers.