API Reference archastro v#0.2.0

Copy Markdown

Modules

Authentication operations generated from x-token-flows.

Rotating bearer credentials returned by login and refresh operations.

Resolved request authorization returned by a token-server provider.

A joined Phoenix Channel owned by an ArchAstro.Socket process.

Phoenix channel for real-time activity feed updates.

Channel for real-time chat messaging.

Channel for real-time custom object collaboration.

Phoenix channel for live task-record updates.

Immutable entry point for the ArchAstro Platform API.

Converts between ArchAstro structs and their JSON wire representations.

Typed REST API or transport error.

Recursive types for values accepted by and decoded from JSON.

A Server-Sent Event.

Typed, lazy stream of decoded server-sent events.

Helpers for session-scoped authentication in LiveView and other multi-user apps.

Supervised Phoenix Channels client backed by Slipstream.

Behaviour and dispatch API for immutable-client credential providers.

Opaque binding between an immutable client and a token-server scope.

GenServer-backed credential provider with a private ETS credential cache.

Optional durable backing store for the default ETS TokenServer.

Terminal event marking the end of a streaming chat completion (SSE done event).

Terminal error event for a streaming chat completion (SSE error event).

The fully assembled assistant message for one run of a streaming chat completion (SSE message_complete event).

Incremental assistant text emitted during a streaming chat completion (SSE message_delta event).

Incremental model reasoning emitted during a streaming chat completion (SSE thinking_delta event).

Incremental tool-call data emitted as the model assembles a tool invocation (SSE tool_call_delta event).

The result of a server-executed tool, streamed back into the run (SSE tool_result event).

The result of an AI chat completion request. Returned by chat completion endpoints after the model finishes generating.

The result returned by an AI image generation or editing operation. Contains the generated image (as inline data or a URL) along with dimension, size, and usage metadata.

A single message in an AI conversation, following the OpenAI-compatible chat format. Used in both request inputs and completion responses.

A tool (function) call emitted by the assistant within an AI message. Mirrors the OpenAI tool-call object format.

The result of executing a tool call, submitted back to the model as a tool-role message. Mirrors the OpenAI tool-result object format.

An access-control list payload that supports either full replacement or targeted patch operations on a resource's grants.

A single access-control grant that pairs a principal with the set of actions it is allowed to perform.

Identifies a principal to be removed from an access-control list.

A single event record in an activity feed, capturing what happened, who caused it, and which resources were involved.

AgentVariant2SourceSolutionCurrentSolution API model.

AgentVariant2SourceSolutionCurrentSolutionOrgLogo API model.

AgentVariant2SourceSolutionCurrentSolutionTemplatesItem API model.

AgentVariant2SourceSolutionCurrentSolutionTemplatesItemDetails API model.

AgentVariant2SourceSolutionCurrentSolutionTemplatesItemDetailsInvokeContract API model.

AgentVariant2SourceSolutionCurrentSolutionTemplatesItemDetailsInvokeContractParticipantsItem API model.

AgentVariant2SourceSolutionCurrentSolutionTemplatesItemDetailsInvokeContractPrefills API model.

AgentVariant2SourceSolutionSolutionOrgLogo API model.

AgentVariant2SourceSolutionSolutionTemplatesItem API model.

AgentVariant2SourceSolutionSolutionTemplatesItemDetails API model.

AgentVariant2SourceSolutionSolutionTemplatesItemDetailsInvokeContract API model.

AgentVariant2SourceSolutionSolutionTemplatesItemDetailsInvokeContractParticipantsItem API model.

AgentVariant2SourceSolutionSolutionTemplatesItemDetailsInvokeContractPrefills API model.

A paginated list of activity feed entries returned by a feed query, with cursors for navigating backward and forward through results.

The entity that authored a message, either a human user or an agent.

An AI agent that can be configured with tools, routines, and skills, and invoked to handle conversations or tasks.

A cloud computer resource provisioned for an agent to use for browser and desktop automation tasks.

A list of agent computers returned by a list query.

The response returned by POST /api/v1/agents. Contains all agent fields plus an optional installed_configs array when a template_bundle was supplied in the request.

An agent environment variable with its secret value masked for safe display in list and show responses.

Flat list of masked environment variables belonging to an agent.

A portable export bundle for an agent, containing everything needed to re-deploy it in another workspace or environment.

Aggregate health profile for an agent, summarizing its current operational status, score, and the full list of setup and health actions.

A single actionable item in an agent's health or setup checklist, carrying the structured data needed to render the item and deep-link to the resolution flow.

Paginated list of agent objects. Use the pagination fields to traverse multiple pages of results.

An agent routine defines a reusable handler — script, preset, or chain — that runs in response to events or on a schedule.

List of agent routine objects belonging to a given agent.

A single execution of an agent routine, capturing its status, inputs, outputs, and timing.

Normalized destination and status for an agent routine run's final-result delivery.

Cursor-paginated list of agent routine run objects, ordered by creation time descending.

A scheduled task created by an agent. Supports one-time and recurring (cron-based) execution patterns.

A durable agent session record representing a single AI task execution. Tracks status, trajectory, result, and any inbox messages delivered to the session.

Paginated list response containing an array of agent session objects.

A skill enabled on an agent, linking the agent to a skill configuration. Controls which capabilities the agent has access to.

Paginated list response containing an array of agent skill objects.

Summary of the Solution and AgentTemplate that an agent was last provisioned from. Returned on single-agent responses; null for hand-built agents and agents whose tracked template or parent Solution has been deleted.

A tool attached to an agent, defining a capability the agent can invoke during a conversation or task run.

Paginated list response containing the tools attached to an agent.

One child-resource change produced by an agent upgrade, describing the action to be taken on a single resource.

One field-level diff entry within an agent upgrade change, describing how a single field will change. baseline and locally_edited are populated only for agent_base entries; child resource entries (tools, routines, skills, computers) carry only field, old, and new.

Response returned by the agent upgrade endpoint, combining the updated agent, its source Solution and template, and the full upgrade diff.

The computed diff and outcome of an agent upgrade operation, including the full list of per-resource changes.

Aggregate counts of each change type produced by an agent upgrade diff.

A versioned artifact produced or managed by an agent, such as a generated file, report, or code output.

A rich attachment associated with a message, such as a file, scraped link, artifact, task, media item, or inline action.

Credential bundle returned after a successful authentication exchange. Contains the access token, refresh token, and the authenticated user.

A named participant slot declared by an automation's workflow. Embedded stages hand work to the agent the invoker names for the slot.

Locked payload and participant values supplied by the automation owner.

A single execution of an automation triggered by a platform event or direct invocation. Captures the run's status, input payload, and final result.

A bug report or freeform feedback submission from any ArchAstro client. Bug reports are write-only for the submitting user and are not returned by any public list or show endpoint.

A single callable tool within a builtin tool catalog entry. Represents one discrete function an agent can invoke.

A catalog entry describing a category of platform-provided (builtin) tools that can be enabled for an agent. Each entry groups one or more individual tools under a shared key, label, and configuration schema.

Empty acknowledgement payload returned by channel message handlers that produce no data. The wire envelope is {"status": "ok", "response": {}}.

List activity feed entries with cursor-based pagination

List activity feed entries with cursor-based pagination

Load additional messages with cursor-based pagination

Load additional messages with cursor-based pagination

Post a new message with optional uploads and reply-to

Post a new message with optional uploads and reply-to

Signal that the current user has started or stopped typing in the thread

Join a team-scoped transient (ephemeral) thread

Join a user-scoped transient (ephemeral) thread

Broadcast when a new message is added to a thread

Broadcast when a message is updated or removed

Broadcast thread-level events (agent updates, read state, unread counts)

Broadcast when a participant (human or agent) starts or stops typing. Ephemeral; never persisted.

Broadcast when the thread's task records change; carries a full snapshot

Response returned after forking a chat thread. Contains the new thread, its initial chat-room snapshot, and the owning team when applicable.

Response returned after loading an additional page of chat messages. Contains a refreshed chat-room snapshot with the newly-fetched messages merged in.

Response returned after marking a chat thread as read. Confirms that the read marker was successfully recorded for the authenticated user.

A participant in a chat thread, which may be either a human user or an AI agent. Exactly one of user or agent is populated depending on type.

Response returned when listing the messages of a joined chat thread. Contains the set of messages currently loaded for the thread.

Response returned after successfully posting a message to a chat thread. Contains the persisted message object echoed back to the sender.

A point-in-time snapshot of a chat room's state, including its loaded messages, member roster, and pagination cursors. Returned when loading or refreshing a thread's message list.

The result of executing a shell command on an agent's computer environment. Contains the captured output and the process exit code.

A versioned config file owned by a team or user, representing a typed artifact such as an agent definition or API tool specification.

Aggregated facet counts for the authenticated viewer's configs. Reports all distinct kinds and path prefixes across the full dataset, regardless of any active list filters.

A facet entry grouping configs by kind, returning the kind name and the number of matching configs visible to the authenticated viewer.

The JSON Schema definition and sample YAML for a specific config kind, used to validate and scaffold new configs of that kind.

A facet entry grouping configs by their leading path segment, returning the prefix and the number of matching configs visible to the authenticated viewer.

A single immutable snapshot of a config's content, created each time the config is saved.

A context document stored within a context source. Carries metadata and size information only; retrieve the full text content via the /content endpoint.

The text content of a context document, optionally sliced by line or byte range. Includes totals and slice boundary fields for the requested unit.

A context ingestion job that processes a context source and populates its documents. Tracks status and timing from submission through completion or failure.

Creation-only private service enrollment response. The raw token is shown once and is omitted from every read schema.

A custom object belonging to an organization. Custom objects store arbitrary structured data defined by a schema type and are scoped to an org, team, or user.

Initial authoritative snapshot returned by a custom-object channel join.

A paginated page of custom objects returned by a list operation. Use the pagination fields to navigate through result sets.

Acknowledges an ephemeral custom-object presence update.

Acknowledges that the current custom-object document reached durable storage.

Response returned after updating one or more fields on a custom object. Confirms the object that was modified and the field values that were applied.

Deployment metadata.

The initial response from an OAuth 2.0 Device Authorization Grant request, containing the codes and URIs needed to complete device authentication.

The result of a completed OAuth 2.0 Device Authorization flow, indicating whether the user approved or denied the device's access request.

An extraction job: yields text from a document or website into a destination namespace, without committing knowledge to an agent.

A produced file tracked by an extraction. Type, size, and URL live on the file it points at.

List response containing agent health actions for a given agent or organization.

Resolved metadata for an image, including its delivery URL, dimensions, and optional references to the underlying storage file or media record.

An installation representing a connection between an agent and an external service or enablement channel. Tracks configuration, lifecycle state, and any bound integration.

A supported installation kind describing a category of external service or enablement channel an agent can be connected to.

List response containing the publicly available installation kinds that can be used when configuring an agent installation.

Paginated list response containing installation objects for an agent.

A source attached to an installation that supplies content for the agent's context. Sources are processed asynchronously after creation.

Paginated list response containing installation source objects attached to an installation.

A slim summary of a single config record created during an agent install transaction. Returned as an entry in AgentCreateResponse.installed_configs.

A minimal, public-safe projection of the user who sent an invite, exposed to unauthenticated recipients so they can render a join screen. Only identity fields are included; sensitive fields such as email address and organization membership are omitted.

A single key-value storage entry belonging to a user. Represents one key/value pair written to a user's isolated storage namespace within an app.

Paginated response envelope for the dual-mode key-value storage list endpoint. End-user (user-JWT) callers receive only data; developer and server-to-server callers also receive pagination metadata fields.

A key-value storage entry enriched with owner information. Developer and server-to-server callers receive user_email and user_name populated; end-user (user-JWT) callers receive those fields as null.

A knowledge source that ingests content into the knowledge base. Sources connect to external systems (e.g. Gmail, GitHub) and continuously or on-demand index items for search.

Describes a single knowledge source kind that can be created through the public API. Use the type value when creating a new knowledge source.

List response containing the knowledge source kinds available for creation via the API.

LLM invocation settings for a routine or chain step. When present, overrides the agent-level model selection.

A processed variant of a media item, such as the original upload or a resized thumbnail, including a signed download URL resolved at request time.

A chat message posted in a thread, including its content, author, attachments, reactions, and optional reply metadata.

Controls visibility and canonical recipient selection for routine-emitted messages.

A compact reaction record embedded in a message's reactions array, representing a single user's reaction to a message.

An inbox notification delivered to a recipient user. Includes type-specific render data resolved at request time.

A single per-channel notification preference for the authenticated viewer, scoped to a notification type and optional app.

The complete set of notification preferences belonging to the authenticated viewer.

A successful OAuth 2.0 token response. Issued by the token endpoint after a completed authorization or device-flow grant.

List activity feed entries query parameters.

DataItemAgentVariant2SourceSolutionCurrentSolutionTemplatesItemDetailsInvokeContractParticipantsItem API model.

DataItemAgentVariant2SourceSolutionCurrentSolutionTemplatesItemDetailsInvokeContractPrefills API model.

DataItemAgentVariant2SourceSolutionSolutionTemplatesItemDetailsInvokeContractParticipantsItem API model.

List installations for an app query parameters.

List agent routine runs query parameters.

List runs for a routine query parameters.

List a routine run journal query parameters.

List agent sessions query parameters.

List agent skills query parameters.

List agent tools query parameters.

List agents query parameters.

List health actions for an agent query parameters.

List working memory entries for an agent query parameters.

Export an agent as an AgentTemplate query parameters.

List schedules for an agent query parameters.

List active workflow work available to the viewer query parameters.

Retrieve raw artifact file content query parameters.

List an automation run journal query parameters.

List configs query parameters.

Retrieve a config query parameters.

Retrieve a config's raw content query parameters.

List a config's version history query parameters.

List config facets query parameters.

List config kinds query parameters.

List system configs query parameters.

List custom objects query parameters.

Retrieve a custom object query parameters.

Fetch an agent avatar image query parameters.

Fetch an org logo image query parameters.

Fetch a publicly shared file query parameters.

List context documents query parameters.

Retrieve a context document's content query parameters.

List knowledge sources query parameters.

List key-value storage entries query parameters.

Retrieve a key-value storage entry query parameters.

List a user's notifications query parameters.

Search organizations query parameters.

List private service enrollments query parameters.

List private services query parameters.

List Slack channel bindings query parameters.

Retrieve a Slack channel binding query parameters.

List solution categories query parameters.

List customer solution instances query parameters.

List solution tags query parameters.

List Solutions query parameters.

Fetch a Solution cover image or gallery screenshot query parameters.

Retrieve a Solution README or asset query parameters.

Retrieve a task query parameters.

List a task's activity query parameters.

List a task's blockers query parameters.

List the tasks a task blocks query parameters.

List comments on a task query parameters.

List a task's subtasks query parameters.

List team memberships query parameters.

List a team's custom objects query parameters.

List task assignees for a team query parameters.

List an owner's tasks query parameters.

Get task activity metrics for a team query parameters.

Search an owner's tasks query parameters.

List threads for a team query parameters.

DataItemParticipatingAgentsItemSourceSolutionCurrentSolutionTemplatesItemDetailsInvokeContract API model.

DataItemParticipatingAgentsItemSourceSolutionCurrentSolutionTemplatesItemDetailsInvokeContractParticipantsItem API model.

DataItemParticipatingAgentsItemSourceSolutionCurrentSolutionTemplatesItemDetailsInvokeContractPrefills API model.

DataItemParticipatingAgentsItemSourceSolutionSolutionTemplatesItemDetailsInvokeContractParticipantsItem API model.

DataItemParticipatingAgentsItemSourceSolutionSolutionTemplatesItemDetailsInvokeContractPrefills API model.

Get threads-created count for a team query parameters.

List replies to a thread message query parameters.

List messages in a thread query parameters.

Retrieve a thread's read status query parameters.

Search messages in a thread query parameters.

List trajectories for a thread query parameters.

List an owner's tasks query parameters.

Search an owner's tasks query parameters.

List threads for a user query parameters.

DataItemParticipatingAgentsItemSourceSolutionCurrentSolutionTemplatesItemDetailsInvokeContract API model.

DataItemParticipatingAgentsItemSourceSolutionCurrentSolutionTemplatesItemDetailsInvokeContractParticipantsItem API model.

DataItemParticipatingAgentsItemSourceSolutionCurrentSolutionTemplatesItemDetailsInvokeContractPrefills API model.

DataItemParticipatingAgentsItemSourceSolutionSolutionTemplatesItemDetailsInvokeContractParticipantsItem API model.

DataItemParticipatingAgentsItemSourceSolutionSolutionTemplatesItemDetailsInvokeContractPrefills API model.

List active workflow work available to the viewer query parameters.

Upgrade an agent from an AgentTemplate

Claim or resume workflow work for an agent

Authenticate with email and password

Register a new user with email and password

Request a magic link for login or registration

Exchange a one-time login token for session tokens

Create a key-value storage entry

Create or update a Slack channel binding

Point a Slack channel's deposit pipe at a staging thread, or turn it off

Import a Solution into the library

Mark a task as blocked by another task

Fail workflow work and route its durable execution

Submit workflow work output and wake its durable execution

Approve a device authorization request

Initiate a device authorization request

Deny a device authorization request

Exchange a grant for OAuth tokens

Create or update a key-value storage entry

Create or update a notification preference

A paginated list of reply messages for a thread. The reply array is returned directly, not nested inside a data wrapper.

Configuration for a preset routine handler. Controls the agent's behavior, session persistence, and model selection for a given routine or chain step.

An immutable private service with complete callable operation contracts.

A private service's durable connector identity. Read responses never contain an enrollment token or certificate.

A secret-free page of private service enrollments.

A documented callable operation exposed by a private service.

A page of private services.

A named preset that defines the execution model and constraints for a routine. Presets are shared definitions; individual routines reference a preset by name.

A forward-paginated journal entry page for an automation or routine run.

An isolated developer sandbox environment used for testing integrations without affecting production data or sending real emails.

An API key scoped to a developer sandbox, used to authenticate requests against sandbox resources.

A binding that connects a Slack channel to an ArchAstro team and one or more agents, enabling those agents to receive and respond to messages in that channel.

Paginated list of Slack channel bindings for the requested integration or team. Use the page and per_page fields to navigate pages of results.

The schema-driven values an installer may lock when provisioning an invoked automation template.

AutomationTemplate-specific details exposed by a Solution template summary.

Paginated list of solution category summaries. Use page and page_size to navigate pages of results.

A solution category that organizes solutions in the catalog, identified by a stable key and optionally nested under a parent category.

A brief representation of an agent that references at least one config bundled by a Solution, included in the dependents preview response.

A preview of the agents and configs that would be affected by deleting a Solution, returned before any deletion occurs so the caller can display a confirmation warning.

A single config entry in a solution upgrade diff, describing what action will be taken on a specific config key.

A reference from another config to an orphaned entry in a solution upgrade diff, explaining why the orphan cannot be safely removed.

Aggregate counts of each action type across all entries in a solution upgrade diff.

The result of importing a Solution bundle into the library, including the Solution config record, a structured import result, and the list of all configs persisted during the transaction.

The machine-readable outcome of a Solution import attempt, indicating whether the import succeeded or requires an upgrade flow to resolve a version conflict.

A non-fatal finding surfaced by a Solution import. The import proceeds despite warnings; validation callers (dry-run) can choose to treat them as failures.

The runtime resource provisioned by installing a Solution, along with a reference back to the source Solution config.

One-time connection details for a webhook-auth Automation install.

A customer-keyed instance stamped from an installed solution template.

A forward cursor-paginated page of customer solution instances.

A paginated collection of Solution summaries, with page metadata for navigating the result set.

A catalog entry for an imported Solution, including its display metadata, bundled templates, owner scopes, and any available upgrade information.

Paginated list of solution tag summaries returned by the list solution tags endpoint.

A single solution tag definition, representing a named classification label that can be applied to solutions.

Template-kind-specific Solution summary details, discriminated by type.

Identity and display metadata for a single template bundled by a Solution, used to represent each wrapped or sibling template at template granularity.

Response returned by the solution upgrade endpoint, containing the solution record, the full upgrade diff, and the resulting installed configs.

The outcome of a solution upgrade operation, including the computed diff and conflict status.

Health check response confirming the API is reachable and indicating whether the caller's token is valid.

A file stored in the platform's object storage, with metadata and a signed URL for downloading its contents.

A task representing a unit of work, optionally assignable to a user or agent.

A comment posted on a task by a user or an agent, including resolved author information.

A task-session lease returned only to its matching holder.

Viewer-safe details about a task's current coding-session lease.

A team within an organization, used to group users and agents and scope resources like configs, agents, and tasks.

A team invite containing a short alphanumeric code that other users can present to join the team.

A record representing a user's or agent's membership in a team, including their resolved identity details and role.

A paginated list of team memberships returned by the list team memberships endpoint.

A chat thread, representing a conversation channel that can be owned by a user, team, or agent and may contain messages, participants, and AI agent activity.

A single message posted to a thread, as seen from the developer portal. Includes sender information, optional attachments, and scoping identifiers.

The read status of a thread for a specific user, indicating how far they have read and how many messages remain unread.

Configuration settings for a thread that control AI agent behavior and other thread-level preferences.

A recorded sequence of AI messages and tool interactions representing a single AI reasoning session. Trajectories are stored as structured message logs and can be replayed or inspected after execution.

Compact summary of an AgentTemplate config referenced by an agent upgrade or source-solution response.

A platform user account. Represents a human or system actor that can own threads, belong to an organization, and interact with the API.

A shareable invite created by a user, optionally scoped to a thread. Recipients can use the invite key to join or start a conversation.

The result of a configuration validation check, indicating whether the config is valid and listing any errors or warnings.

Execution state of the background worker processing a routine run. Reflects the current job status and retry progress.

Summary of the durable workflow execution journal associated with a run.

One ordered, replayable record from a durable workflow journal.

Externally executable work yielded by a durable workflow.

Result of polling an agent's durable workflow work queue.

A claimed workflow work item and its caller-held lease token.

Active durable workflow work available to the viewer.

A key-value memory record stored for an agent, optionally scoped to a user. Memory entries persist across invocations and may carry an expiration time.

Paginated list of working memory entries stored for an agent. Includes page metadata to support sequential page traversal.

Sentinel used to distinguish an omitted JSON field from an explicit null.

ActivityFeed API resource.

AgentComputer API resource.

AgentEnvVar API resource.

AgentHealthAction API resource.

AgentInstallation API resource.

AgentInstallationInstallationSource API resource.

AgentRoutineRun API resource.

AgentRoutine API resource.

AgentRoutineRun API resource.

AgentSession API resource.

AgentSkill API resource.

AgentTool API resource.

Agent API resource.

AgentAgentComputer API resource.

AgentAgentEnvVar API resource.

AgentAgentInstallation API resource.

AgentAgentTool API resource.

AgentWorkingMemory API resource.

Schedule API resource.

AgentWorkItem API resource.

Chat API resource.

Completion API resource.

Embedding API resource.

Image API resource.

Artifact API resource.

AutomationRun API resource.

Automation API resource.

BugReport API resource.

Config API resource.

Kind API resource.

System API resource.

CustomObject API resource.

Extraction API resource.

File API resource.

InstallationSource API resource.

Invite API resource.

KnowledgeDocument API resource.

KnowledgeIngestion API resource.

KnowledgeSource API resource.

KnowledgeSourceKind API resource.

Kv API resource.

NotificationPreference API resource.

Notification API resource.

Oauth API resource.

Device API resource.

Org API resource.

PrivateServiceDefinition API resource.

PrivateServiceEnrollment API resource.

PrivateService API resource.

Sandbox API resource.

SlackChannelBinding API resource.

SolutionCategory API resource.

SolutionInstance API resource.

SolutionTag API resource.

Solution API resource.

Statu API resource.

Task API resource.

Blocker API resource.

Comment API resource.

Lease API resource.

TeamMembership API resource.

Team API resource.

TeamCustomObject API resource.

Member API resource.

TeamTask API resource.

TeamThread API resource.

ThreadMessage API resource.

Reaction API resource.

Thread API resource.

ThreadMember API resource.

Setting API resource.

Tag API resource.

Trajectory API resource.

User API resource.

UserTask API resource.

UserThread API resource.

WorkItem API resource.