Every call you make in MCPFlo — whether invoking a tool, reading a resource, or rendering a prompt — is recorded in a lightweight history, so you can look back at what you ran without redoing the work from scratch.

Per-capability history

Each individual tool, resource, or prompt keeps its own history, capped at the 50 most recent calls. Selecting a capability shows its history alongside its invocation form, so you can see past runs in context without leaving the view.

The “All” tab

Alongside per-capability history, a unified All tab aggregates every call across every connected server into a single, cross-server activity log for the current session — useful for reviewing everything you’ve done recently, regardless of which capability or server it belonged to.

What’s recorded

For each history entry:

• The arguments/parameters used for the call
• The full response (or error) that came back
• Enough context to identify which server and capability it belongs to (relevant in the All tab)

Non-persistent by design

History is cleared when the app closes — it does not persist to disk across sessions. This is a deliberate limitation (see README’s “Not supported” list): history is meant as a working scratchpad for your current testing session, not a permanent audit log.

If you need to preserve a specific result long-term, copy it out of the raw JSON-RPC view (see Understanding Responses) rather than relying on history to still be there next time you open the app.

Related

• See Replaying Past Calls for re-running a past call directly from history.
• See Reading Resources & Rendering Prompts for what gets recorded when the capability isn’t a tool.