Agents

Agent data

Nothing agent-specific is compiled into the any server. The harness declares its stores as runtime datasets on objects it derives itself, owns the record shapes and validation, and reads and writes them through the generic query, modify and upsert surface. You can do the same.

Where it lives

At serve boot anyrt registers the bao/v1 bundle in the agent space and derives one child per store — deterministic ids every device computes offline, so two devices never mint two brains.

Turn logs are per chat: the general chat is the general-chat/v1 bundle root, and its log is a child of that bundle, so it cascade-deletes with the chat. Each store is ensured by its writer, idempotently, on first use.

Resolve a child yourself:

curl -X POST http://127.0.0.1:7001/v1/spaces/$SPACE/bundles/bao%2Fv1/children \
  -H 'content-type: application/json' \
  -d '{"seed": "bao/brain/v1"}'

Datasets

agent_turns (idRule: user, id = zero-padded seq, all fields write-once, deleteBy: author)

agent_chunks — seq, level (1 over turns, 2+ over chunks), fromSeq/toSeq, summary, periodStart/periodEnd, unitsCovered; indexed under scope history.

agent_memory_items (idRule: auto, deleteBy: author) — write-once category, context (both required), validFrom, chatId, fromAgent; author-mutable body, tags, entities, keywords, confidence, importance, salience, accessCount, edges; stamps creator, createdAt, modifiedAt; search {title: context, text: body, scope: "agent"}. agent_job_state holds cron cursors (one record per job id); agent_roi_injections is auto-recall's injection log.

agent_config (idRule: user, id = the dotted key, dynamic: true) — key, secret (boolean), value (space-wide, synced), localValue (scope: local, this device only). Resolution is localValue ?? value ?? default; defaults such as llm.tier.* and search.provider.* ship inside anyrt.

agent_secrets (idRule: user, id = the secret ref) — key, secret, and value with scope: local: the plaintext never syncs and guest reads are refused. See Connectors.

agent_triggers / agent_trigger_runs (idRule: user, dynamic: true, zero declared fields) — the whole definition (name, kind cron/once/event, spec, program, args, owner, enabled, limits, maxConsecutiveFailures) plus the rollup (lastRunAt, lastStatus, lastDurationMs, runCount, lastRunRef) rides the free keyspace. Contract in Scheduling.

Reading it yourself

The last five turns of the general chat:

curl -X POST http://127.0.0.1:7001/v1/spaces/$SPACE/query \
  -H 'content-type: application/json' \
  -d '{"objectId": "'$LOG'", "dataset": "agent_turns",
       "sort": ["-seq"], "limit": 5}'

any query $SPACE $LOG agent_turns --sort -seq --limit 5

Every memory item in one category, live:

curl -N -X POST http://127.0.0.1:7001/v1/spaces/$SPACE/query/subscribe \
  -H 'content-type: application/json' \
  -d '{"objectId": "'$BRAIN'", "dataset": "agent_memory_items",
       "filter": {"category": "decision"}, "sort": ["-createdAt"], "limit": 50}'

Memory and history also participate in search under scopes agent and history.

Scheduling a job from the agent

Because triggers are records, the agent creates one by writing a record — a reminder, for example:

c = use("agent:any@v1")
c.upsert_records(space, triggers_id, "agent_triggers", [{
  "id": "remind-standup",
  "fields": {"kind": "once", "spec": {"at": 1756112400},
             "program": "agent:remind@v1",
             "args": {"space": space_id, "chatId": chat_id,
                      "text": "standup in 10 minutes"},
             "enabled": True}}])

The running device adopts it on the next tick; a once fires when now >= at, then auto-disables and keeps its record as the audit trail. See Once.

Why it matters. The agent's whole operational state is a handful of datasets in your encrypted space: queryable, subscribable, exportable, deletable, and synced to your other devices — with no schema you cannot read and no store you cannot leave.