Getting startedSet up integrationsCreate your first initiativeInvite your teamPlan today's workShare your first update
OverviewQuickstartRuntime manifestACP contractDocuments & toolsMCP toolsDistributeRun & monitorBuild cloud agentAgent sessionsReferenceLimits & securityTroubleshooting
DocsAPI Reference

Main

  • Home
  • About
  • Pricing
  • Vault
  • Changelog
  • Docs

Features

  • Roadmaps
  • Planning
  • Standups
  • Status updates
  • Insights
  • AI assistant / MCP
  • Integrations

Solutions

  • Startups
  • Dev shops / agencies
  • Software teams
  • Internal IT & platform teams

Alternatives

  • vs Jira
  • vs Linear
  • vs Asana
  • vs Monday.com
  • vs ClickUp
  • vs Notion

Company

  • Blog
  • Security
  • Log in
  • Sign up
  • Terms of Use
  • Privacy Policy

Resources

  • Docs
  • Community
  • API reference
  • CLI
  • Desktop app
  • SDK

© 2026 One Horizon. All rights reserved

FacebookInstagramThreadsXRedditTikTokYouTubeMedium


Limits and security

Custom runtimes run on machines One Horizon does not control, so every request they make crosses a trust boundary. The API gateway validates each request body against the published schema before it reaches any backend service, and rejects oversized or malformed input with 400. This page lists the restrictions that apply to agent traffic. All values are enforced server-side and mirrored in the API reference.

Authentication and scope

  • Agent lifecycle endpoints (register, claim, patch, complete, fail, release, heartbeat, activities, control signals, sessions) require an OAuth user token. Workspace API keys are rejected on these operations.
  • Every request is scoped to one workspace. API-key actors may only reach their own workspace; a mismatch returns 403.
  • The per-turn MCP token is scoped to the session's task, workflow run, and declared artifacts, and expires when the turn ends. Tool calls beyond the step's declared outputs fail. See documents and tools.
  • Your agent never receives the user's OAuth token, workspace API keys, or other sessions' data.

Rate limits

Requests are metered per actor with a token-bucket limiter. Exceeding a bucket returns 429 with a retry in N seconds message.

TrafficPolicy
Interactive OAuth requestsStandard per-user window with a short burst allowance
Agent control-plane (poll, heartbeat, claim, patch, activities, control signals)Higher automated window sized for continuous worker traffic
Failed authenticationPer-credential lockout window
API key creationPer-workspace hourly cap, plus a maximum number of keys per workspace

List endpoints are paginated: agent and session lists cap at 100 items per page, document lists at 200.

What an agent sends into the platform

Free-text and identifier fields are length-limited so a runtime cannot flood the database or the task timeline.

FieldEndpointMaximum
Session planpatch / complete / fail session20,000 characters
Error messagepatch / complete / fail / cancel session20,000 characters
Activity messagecreate activity50,000 characters
Activity plancreate activity20,000 characters
External URLactivity, session update2,048 characters
Provider session / turn / run IDsession update512 characters
Working directory, worktree, transcript pathsession update4,096 characters
Model, model providersession update256 characters
Workflow branch provider / repo / branchsession update128–512 characters
Claim defer reasondefer claim2,000 characters
Resource snapshotdefer claim20,000 characters
Task comment bodysave comment50,000 characters
Comment source / display name / actor IDsave comment64 / 256 / 256 characters
Document title / bodysave document512 / 1,048,576 characters
Agent and worker custom instructionsagent / worker profile2,000 characters

Custom instructions are trusted guidance shown to the runtime. Do not store secrets, tokens, or credentials in them.

What the platform sends to an agent

FieldWhereMaximum
Session promptqueue session50,000 characters
Resume inputresume session20,000 characters

Prompt content, task title, description, comments, and resume input arrive as untrusted context. Trusted platform policy and workflow instructions are marked separately. Your agent must not let untrusted content override trusted policy — see the trust boundaries in the ACP contract.

Uploads and artifacts

UploadAcceptedMaximumValidation
File (screenshots, recordings)image/* except image/svg+xml, and video/*100 MBContent type checked; pre-signed PUT sent from a server, not a browser
Agent iconSVG only64 KB decodedSent as base64 and validated synchronously against a markup allowlist; scripts, event handlers, and <image> elements (embedded or referenced raster content) are rejected before storage
Media download (proxy)image and video types100 MBContent type is sniffed against the declared type; SVG is refused; served as a no-store attachment with X-Content-Type-Options: nosniff
Runtime source artifactarchive from an allowlisted source100 MBFetched over HTTPS only, redirects restricted to allowlisted hosts, verified against a SHA-256 digest

Stored files are private. The assetUrl returned by an upload is a proxied URL gated by workspace authentication, not a public link. Verification media left in ONE_WORKER_ARTIFACTS_DIR is uploaded by the host after the turn under the same rules.

The general file-upload endpoint accepts opaque images and video only. It rejects SVG because markup can carry a script payload; upload SVG icons through the dedicated agent-icon endpoint, which inspects the markup first.


PreviousReferenceNextTroubleshooting

Reference

Reference runtime responsibilities, workflow context, artifact documents, environment variables, and API operation families.

MCP tools

Every MCP tool the worker runtime can register for a turn, generated straight from the runtime tool catalog.

Workflows

Create, publish, launch, and operate workflows with agent steps and human review gates.

Overview

Plug any ACP-compatible agent into the worker runtime. One Horizon handles auth, sessions, tools, and monitoring; you build the agent.

  • Authentication and scope
  • Rate limits
  • What an agent sends into the platform
  • What the platform sends to an agent
  • Uploads and artifacts
  • Back to top