# List sessions URL: https://docs.hanzo.ai/docs/openapi/agent/get-agent-sessions > Returns the caller org's live sessions, newest first — each with its event count, its direct-child count and a one-line preview of its latest event. `GET /v1/agent/sessions` > [Agent →](https://docs.hanzo.ai/docs/openapi/agent) · [All API references →](https://docs.hanzo.ai/docs/openapi) · [Get started →](https://docs.hanzo.ai/docs/guides) | | | | ------------- | ---------------------------------------- | | **Address** | `https://api.hanzo.ai/v1/agent/sessions` | | **Method** | `GET` | | **Operation** | `get_agent_sessions` | | **Auth** | `Authorization: Bearer $HANZO_API_KEY` | Returns the caller org's live sessions, newest first — each with its event count, its direct-child count and a one-line preview of its latest event. With no filter it returns ROOT sessions only, so a dashboard shows one row per flow rather than one per subagent; ?root= or ?parent= descends. ## Request 8 fields. | Field | In | Type | Required | Description | | --------- | ----- | ------- | -------- | ------------------------------------------------------------------------------------------------------------------------------------------------ | | `root` | query | string | — | Root scopes the page to one subagent tree (its root session id). | | `parent` | query | string | — | Parent scopes the page to the direct children of one session. | | `status` | query | string | — | Status filters to running, paused, done or error. | | `project` | query | string | — | Project filters to the sessions tagged with one product slug. | | `room` | query | string | — | Room filters to the sessions started in one collaborative room — the query a space view runs to show what has been run in it. | | `kind` | query | string | — | Kind filters to the sessions of one kind of run: "coding" lists coding runs, each carrying its repo, base, branch, environment and pull request. | | `limit` | query | integer | — | Limit caps the page. | | `after` | query | string | — | After is the `next` of the previous page. | ## Response | Status | Body | Meaning | | --------- | ----------------- | ------- | | `200` | agent.sessionList | ok | | `default` | problem-details | refused | `200` body — 47 fields. | Field | In | Type | Always | Description | | ------------------------------- | ---- | --------------------- | ------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `next` | body | string | — | Next is the cursor for the page after this one; empty on the last page. | | `sessions` | body | agent.sessionView\[] | — | Sessions is the matching sessions, each with its event and child counts and a one-line preview of its latest event. | | `sessions[].account` | body | string | — | Account is which subscription or API account under that provider served it. | | `sessions[].actor` | body | string | — | Actor is WHO this session belongs to, as "org/sub" — the same identity a run is billed under. | | `sessions[].agent` | body | string | — | Agent is the label the surface running this session calls itself by ("hanzo-dev"), up to 128 characters. | | `sessions[].base` | body | string | — | Base is the branch the run started from. | | `sessions[].branch` | body | string | — | Branch is the branch the run pushes its work to, and the only one it may write. | | `sessions[].children` | body | integer (int64) | — | Children is the DIRECT fan-out — how many sessions name this one as parent — and not the size of the subtree. | | `sessions[].createdAt` | body | string | — | CreatedAt is when the row was written, same format. | | `sessions[].cwd` | body | string | — | Cwd is the directory the session is working in NOW, not the one it started in: a linked shell moves around, and a card showing where `hanzo link` was run answers "which work is this" with something that was true once. | | `sessions[].endedAt` | body | string | — | EndedAt is when it reached done or error, same format. | | `sessions[].environment` | body | string | — | Environment is where the run executes: "sandbox" for a Hanzo sandbox, or the id of the org's machine it was routed to (the same id Target carries). | | `sessions[].events` | body | integer (int64) | — | Events is how many turns the session's log holds, counted at read time. | | `sessions[].host` | body | string | — | Execution context (mission-control): the machine/repo/cwd a card shows and the run-target a session is dispatched to. | | `sessions[].id` | body | string | — | ID is the session's handle, minted here as "sess\_" + 32 hex characters. | | `sessions[].kind` | body | string | — | Kind is what sort of run this session is the view of — "coding" for a coding run. | | `sessions[].lastEvent` | body | agent.lastEventView | — | | | `sessions[].lastEvent.actor` | body | string | — | Actor is who produced the turn, defaulted to the calling principal when the writer named nobody. | | `sessions[].lastEvent.at` | body | string | — | At is when the turn was recorded, RFC 3339 in UTC to the second. | | `sessions[].lastEvent.kind` | body | string | — | Kind is what the turn was, from the log's closed six: message, tool-call, spawn, log, status, control. | | `sessions[].lastEvent.preview` | body | string | — | Preview is the first 240 bytes of the event's payload, cut without regard for the JSON inside it — it is a string to SHOW, never a value to parse. | | `sessions[].lastEvent.seq` | body | integer (int64) | — | Seq is that event's position in the session's log — monotonic from 1, per session. | | `sessions[].mode` | body | string | — | Mode is what the run may do to its repository: "build" edits, commits and pushes its branch; "plan" reads and answers with a plan, writing nothing. | | `sessions[].org` | body | string | — | Org is the caller's OWN tenant, echoed so a client can build the public build URL (/builds/:org/:project) without a second call or a guess. | | `sessions[].parentSessionId` | body | string | — | ParentSessionID is the session that spawned this one, making this a subagent of it. | | `sessions[].pr` | body | string | — | PR is the address of the pull request the run proposed. | | `sessions[].progress` | body | agent.sessionProgress | — | | | `sessions[].progress.activity` | body | string | — | Activity is the one line saying what the run is doing right now ("running the reaper's tests"), up to 120 characters, in the model's words. | | `sessions[].progress.at` | body | string | — | At is when this was determined, RFC 3339 in UTC to the second. | | `sessions[].progress.estimated` | body | boolean | — | Estimated says a MODEL produced this, from the run's transcript, and it may be wrong. | | `sessions[].progress.pct` | body | integer (int64) | — | Pct is how much of the run is done, 0 to 100. | | `sessions[].progress.phase` | body | string | — | Phase is what shape the run is in: running, blocked, done, error, or unknown when nothing has estimated it yet. | | `sessions[].project` | body | string | — | The readable build: the product this session built and whether its story is public (provenance.go). | | `sessions[].provider` | body | string | — | Provider is the linked AI account's provider (claude \| codex \| hanzo \| …) that served this run. | | `sessions[].published` | body | boolean | — | Published is the author's decision to let anyone read this session's story at the public build route. | | `sessions[].repo` | body | string | — | Repo is the code the session is working on, as the surface reported it. | | `sessions[].room` | body | string | — | Room is the collaborative room this run was started in (HIP-0523), empty when it came from anywhere else — a CLI, a schedule, an API call. | | `sessions[].rootSessionId` | body | string | — | RootSessionID is the top of this session's tree, inherited from the parent and shared by every node in one flow. | | `sessions[].sandbox` | body | string | — | Sandbox is the id of the sandbox the run works in ("m\_" and 24 hex), the handle its terminal and its screen are opened with (/v1/sandbox/{id}/terminal, /screen). | | `sessions[].startedAt` | body | string | — | StartedAt is when the session opened, RFC 3339 in UTC to the second. | | `sessions[].status` | body | string | — | Status is one of exactly four: running, paused, done, error. | | `sessions[].target` | body | string | — | Target is the registered run-target this session is dispatched to — a machine the org claimed, resolved same-org when it was set, so it can never point at another tenant's computer. | | `sessions[].taskRunId` | body | string | — | TaskRunID is that workflow's particular run. | | `sessions[].taskWorkflowId` | body | string | — | TaskWorkflowID is the hanzoai/tasks durable workflow that actually EXECUTES this session — this registry is the view, control and stream layer over it. | | `sessions[].terminal` | body | string | — | Terminal is where this session can be WATCHED — the URL the machine published for its live terminal. | | `sessions[].title` | body | string | — | Title is the human line a card shows ("ship the landing page"), up to 512 characters. | | `sessions[].updatedAt` | body | string | — | UpdatedAt is the session's last-activity clock, same format. | Failure carries the platform error shape — see [Errors](https://docs.hanzo.ai/docs/errors). ## Examples **CLI** ```bash hanzo agent sessions list ``` **SDK** **.ts** ```ts import { Configuration, AgentApi } from 'hanzoai'; const api = new AgentApi(new Configuration({ accessToken: process.env.HANZO_API_KEY })); const { data } = await api.getAgentSessions(); ``` **.py** ```python from hanzoai.cloud import ApiClient, Configuration from hanzoai.cloud.api import AgentApi client = ApiClient(Configuration(access_token=os.environ["HANZO_API_KEY"])) result = AgentApi(client).get_agent_sessions() ``` **.go** ```go cfg := hanzoai.NewConfiguration() cfg.AddDefaultHeader("Authorization", "Bearer "+os.Getenv("HANZO_API_KEY")) client := hanzoai.NewAPIClient(cfg) resp, _, err := client.AgentAPI.GetAgentSessions(context.Background()).Execute() if err != nil { return err } ``` **.rs** ```rust use hanzo_client::apis::{configuration::Configuration, agent_api}; let mut cfg = Configuration::new(); cfg.bearer_access_token = std::env::var("HANZO_API_KEY").ok(); let result = agent_api::get_agent_sessions(&cfg, Default::default()).await?; ``` **.java** ```java import ai.hanzo.cloud.ApiClient; import ai.hanzo.cloud.api.AgentApi; ApiClient client = new ApiClient(); client.setBearerToken(System.getenv("HANZO_API_KEY")); var result = new AgentApi(client).getAgentSessions(); ``` **HTTP** ```bash curl https://api.hanzo.ai/v1/agent/sessions \ -H "Authorization: Bearer $HANZO_API_KEY" ``` **MCP** MCP reaches **agent** through the `agent` tool, which names its 38 operations with its own verbs — this one among them, under a name only MCP declares. `describe` explains any of them: ```bash curl -X POST https://api.hanzo.ai/v1/mcp \ -H "Content-Type: application/json" \ -d '{ "jsonrpc": "2.0", "id": 1, "method": "tools/call", "params": { "name": "describe", "arguments": { "op": "list_agents" } } }' ``` --- [Agent API](https://docs.hanzo.ai/docs/openapi/agent) · [All Hanzo APIs](https://docs.hanzo.ai/docs/openapi) · [Interactive reference](https://docs.hanzo.ai/docs/openapi/reference)