PromptLayer + Memtro

Only when you ask, and only what it has been granted: publish a new version of a prompt (replace its messages or text, optionally move release labels to it); create a new prompt template; create a release label on a prompt version; move a release label to another version (changes what code fetching that label gets); delete a release label; score a logged request (0 to 100); add metadata to a logged request; create a folder; run an agent workflow (spends LLM provider credits); run an evaluation (spends LLM provider credits); rename an evaluation; archive an evaluation; add rows to a table sheet; delete a table. Every other tool reads, searches or lists. The Connections page labels the tools that act, an administrator grants each one by name to the teams that may use it, and a personal connection acts only after its owner allows it.

A personal connection serves only your own requests. A shared connection serves the organisation: everyone, or only the teams an admin chooses. Admins can always add shared connections; other members need the permission.

Credentials are encrypted at rest. Tool results are cached briefly so agent loops do not repeat identical calls, and the facts the model extracts are saved to your organisation's memory with provenance (system, record and link) so they can be checked and superseded.

API key. In PromptLayer open Settings, then API keys, and create a key. Each workspace has its own keys.

Yes. Turn events on in the connection's Events tab on the Connections page, then add the address it shows as a webhook in PromptLayer, choosing the same kinds; the tab shows PromptLayer's own steps. A job set to start "When an event arrives" then runs once for each matching event, with the event as its input. The run treats the event as outside data, never as instructions, and anything the run remembers is kept as unverified.