Tag

#agents

6 pages tagged #agents.


Architecture architecture/ 1 page

Agent surfaceReferenceDraft

The MCP tools agents call, and the session bootstrap convention.

architecture/agent-surface #architecture #mcp #tools

Decisions adr/ 4 pages

Every agent has an identity and a trust level; trusted reads all, untrusted is confined to its spaces.

adr/0012-agent-identity-and-trust #adr #security #auth

Why an agent posting a question or an approval may say how long it can wait, why the hub then resolves that one item as itself, and why this is not retention.

adr/0027-deadlines-on-open-items #adr #inbox #approvals #questions

Why the hub carries what needs an agent's attention on the next tool result it makes, and offers standing subscriptions, instead of a push channel or a per-harness poll.

adr/0024-notifications-on-the-tool-result #adr #mcp #notifications

An agent cannot leave unbounded open items on the human; the writer refuses past a configurable cap.

adr/0017-inbox-action-item-cap #adr #inbox #safety

Usage usage/ 1 page

Wire a harness to the hub, start sessions, read the feed and the recovery handoff, and bring existing notes over.

usage/agents #usage #mcp #sessions #migration #config


All tags