AI agents use store_credential to create or update resources in CHARLIE — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your CHARLIE environment.
The tool writes (stores) a credential securely. It does not delete, execute code, or move money. However, storing credentials carries high severity because a misuse — e.g., storing a malicious or incorrect credential, or associating the wrong secret with a service — could silently compromise authentication flows or exfiltrate secrets if the encryption/storage is later accessed by other tools (e.g., get_credential).
From the tool's definition "Encrypt and store a credential" — creates/persists a new encrypted secret; "NEVER returns the value" confirms write-only storage with no read-back.
Attacks that exploit this kind of access
Encrypt and store a credential. NEVER returns the value. It is categorised as a Write tool in the CHARLIE MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the CHARLIE MCP server in PolicyLayer and add a rule for store_credential: 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 CHARLIE. Nothing to install.
store_credential 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 store_credential 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 store_credential. 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.
store_credential is provided by the CHARLIE MCP server (t3cch/charlie). 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 →