APIs
Generators
Stacks
Projects

Docs
History
Schema
Generate

Filter operations

AdminGate2

Agents62

post/agents/webhook/{webhook_id}Public webhook endpoint. Authenticated via `X-Webhook-Token` header.
get/projects/{project_id}/agents
post/projects/{project_id}/agents
get/projects/{project_id}/agents/cli-status
get/projects/{project_id}/agents/runs
get/projects/{project_id}/agents/runs/latest-for-source
get/projects/{project_id}/agents/runs/{run_id}
post/projects/{project_id}/agents/runs/{run_id}/cancel
post/projects/{project_id}/agents/runs/{run_id}/retryRetry a completed, failed, cancelled, or no_fix run with the same trigger context. Creates a new run record and spawns the executor.
get/projects/{project_id}/agents/runs/{run_id}/streamSSE endpoint for real-time streaming of run events. Polls the agent_run_logs table every 500ms for new entries and streams them. Closes when the run reaches a terminal status.
get/projects/{project_id}/agents/sandbox-status
post/projects/{project_id}/agents/smoke-testRun a smoke test to verify the selected AI CLI works in the environment where agents will actually execute (host or sandbox container). If no `provider_id` is supplied the globally active provider is tested.
get/projects/{project_id}/agents/{slug}
put/projects/{project_id}/agents/{slug}
delete/projects/{project_id}/agents/{slug}
get/projects/{project_id}/agents/{slug}/runs
post/projects/{project_id}/agents/{slug}/trigger
get/projects/{project_id}/autofixer/runs/{run_id}/streamSSE endpoint: streams run log events in real-time. Polls every 500 ms. Keeps the connection open through "analyzed" and "fix_ready" waiting states; closes only on terminal statuses.
get/projects/{project_id}/mcp-servers
post/projects/{project_id}/mcp-servers
get/projects/{project_id}/mcp-servers/{slug}
put/projects/{project_id}/mcp-servers/{slug}
delete/projects/{project_id}/mcp-servers/{slug}
get/projects/{project_id}/mcp-servers/{slug}/config/{field}
get/projects/{project_id}/skills
post/projects/{project_id}/skills
post/projects/{project_id}/skills/uploadUpload a skill with an archive (tar.gz) — project-scoped.
get/projects/{project_id}/skills/{slug}
put/projects/{project_id}/skills/{slug}
delete/projects/{project_id}/skills/{slug}
get/projects/{project_id}/skills/{slug}/archiveDownload a skill's archive (tar.gz) — project-scoped.
post/settings/agent-tokenSave an encrypted AI provider token for use in sandbox containers.
get/settings/ai-providersList the AI provider catalog. Includes per-provider "is a credential configured?" so the settings UI can render configured/not-configured badges without leaking the encrypted credential.
patch/settings/ai-providers/{provider_id}Update provider-scoped settings without touching the saved credential. Today that means just `default_model`; future per-provider settings (base URL overrides, request headers, etc.) can land here too without changing the shape of `save_credential`.
post/settings/ai-providers/{provider_id}/activateActivate a provider as the platform-wide default. Refuses to activate a provider that doesn't have a credential saved yet — the UI enforces the same rule on the button, but we re-check server-side so a stale tab can't bypass it.
post/settings/ai-providers/{provider_id}/credentialSave (or replace) a provider's credential. The credential is encrypted with `EncryptionService` and stored inside `agent_sandbox.providers[provider_id].credentials_encrypted`.
post/settings/ai-providers/{provider_id}/credential/import-localImport the credential already used by this provider's CLI on the Temps host. Discovery and encryption happen entirely server-side; the plaintext credential is never serialized into the response or browser state.
post/settings/ai-providers/{provider_id}/credential/verify-savedVerify a previously imported OpenCode credential against a user-selected model without asking the browser to receive or resubmit the secret.
get/settings/ai-providers/{provider_id}/harness-updateInstalled and latest version of the provider's harness CLI in your persistent workspace, and whether an update is available.
post/settings/ai-providers/{provider_id}/harness-updateInstall the latest release of the provider's harness CLI in your persistent workspace so newly released models become available.
get/settings/ai-providers/{provider_id}/loginCurrent sign-in state for a provider: phase, link, device code, elapsed time.
post/settings/ai-providers/{provider_id}/loginStart signing an AI provider's CLI in to a subscription.
post/settings/ai-providers/{provider_id}/login/cancelCancel an in-progress sign-in and stop its CLI process.
post/settings/ai-providers/{provider_id}/login/codeHand the provider's authorization code to the waiting CLI.
post/settings/ai-providers/{provider_id}/models/refreshRefresh exactly one provider's account-aware model inventory. This is an explicit operation because it starts a CLI process (and, for Claude workspaces, creates or wakes the user's persistent sandbox). Authorization therefore matches the provider's chat execution requirement instead of requiring unrelated settings mutation access. The AI service applies a provider-scoped single-flight and cooldown before doing that work.
post/settings/ai-providers/{provider_id}/preflight
post/settings/ai-providers/{provider_id}/smoke
get/settings/mcp-servers
post/settings/mcp-servers
get/settings/mcp-servers/{slug}
put/settings/mcp-servers/{slug}
delete/settings/mcp-servers/{slug}
get/settings/mcp-servers/{slug}/config/{field}
post/settings/sandbox-rebuild
get/settings/sandbox-status
get/settings/skills
post/settings/skills
post/settings/skills/uploadUpload a skill with an archive (tar.gz) — global.
get/settings/skills/{slug}
put/settings/skills/{slug}
delete/settings/skills/{slug}
get/settings/skills/{slug}/archiveDownload a skill's archive (tar.gz) — global.

AI Applications30

AI Chat44

AI Gateway14

AI Gateway Admin12

AI Gateway Limits5

AI Gateway Pricing1

AI Gateway Usage8

AI Provider Status3

Alarms14

Alerts6

Analytics43

Analytics Ingest Keys6

API Keys9

Audit Logs2

Authentication29

Autofixer7

Backups36

Blob6

Blob Management6

CI builder3

CI runner6

Cloud8

Compose3

Containers8

Crons3

Custom Domains8

Dashboards5

Deployment Tokens6

Deployments25

DNS1

DNS Providers14

DNS Records4

Domains19

Email Domains12

Email Providers9

Email Tracking7

Email Validation1

Emails5

Environments2

error-alert-rules5

error-tracking10

Events13

External Images7

External Services50

External Services - Query11

Feature Flags9

Files1

Funnels7

GenAI2

geo2

Git Provider Connections3

Git Providers35

HTTP Checks12

Imports5

Insights1

Internal DNS2

IP Access Control6

KV Management6

KV Store7

Load Balancer5

Logs11

Metrics17

Nodes28

Notification Preferences4

Notification Providers15

Notification Routes7

Observability2

Operations1

OTel15

OTel Facets4

OTel Ingest6

Performance6

Platform6

Postgres Upgrades7

Presets2

Preview Gateway6

Project Groups22

Project IP Access8

Projects47

Proxy Logs14

Public Repositories6

Repositories6

Restore6

Revenue16

Runtime health3

Sandboxes40

Secrets7

sentry-compat6

sentry-ingestor3

Settings16

Snapshots5

source-maps8

Static Bundles4

Status Page32

Teams13

Telemetry2

Telemetry Logs1

Telemetry Metrics4

Templates7

Traces8

Traefik Discovery6

Traffic Delivery11

Users13

Visitor Activity5

Vulnerability Scans8

Webhook Deliveries3

Webhooks6

Workflows1

get/projects/{project_id}/dsnsList all DSNs for a project
post/projects/{project_id}/dsnsCreate a new DSN for a project
post/projects/{project_id}/dsns/get-or-createGet or create DSN for a project/environment/deployment combination
post/projects/{project_id}/dsns/{dsn_id}/regenerateRegenerate DSN keys (rotate keys)
post/projects/{project_id}/dsns/{dsn_id}/revokeRevoke (deactivate) a DSN

Agents

Changes to Hand the provider's authorization code to the waiting CLI.

post/settings/ai-providers/{provider_id}/login/code

Oct 13, 202622e058697ad7 Revision changes
- info
  endpoint added