Cache which .skill.md / SKILL.md file was used for a task so similar future tasks route to the same skill without re-choosing. Record outcome to let recall_skill_choice rank successful choices higher.
AI agents use remember_skill_choice to create or update resources in GraphHub — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your GraphHub environment.
This tool writes/stores a cached record of skill file choices and their outcomes into persistent session memory. It creates or modifies data (the skill routing cache) in a reversible way. No code is executed, no data is deleted, and no financial transactions occur. The blast radius is low — misuse could only pollute the skill routing cache with incorrect associations.
From the tool's definition Cache which .skill.md / SKILL.md file was used for a task... Record outcome to let recall_skill_choice rank successful choices higher.
Documented attack patterns abuse exactly the kind of access remember_skill_choice gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and GraphHub, and nothing reaches the server without passing your rules. This is the rule we recommend for remember_skill_choice:
{
"version": "1",
"default": "deny",
"tools": {
"remember_skill_choice": {
"limits": [
{
"counter": "remember_skill_choice_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}remember_skill_choice 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.
Cache which .skill.md / SKILL.md file was used for a task so similar future tasks route to the same skill without re-choosing. Record outcome to let recall_skill_choice rank successful choices higher. It is categorised as a Write tool in the GraphHub MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the GraphHub MCP server in PolicyLayer and add a rule for remember_skill_choice: 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 GraphHub. Nothing to install.
remember_skill_choice 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 remember_skill_choice 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 remember_skill_choice. 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.
remember_skill_choice is provided by the GraphHub MCP server (slnquangtran/graph-hub). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from GraphHub, 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.
32 GraphHub tools catalogued and risk-classified — across an index of 43,000+ MCP servers.