Run, manage and build your missions from your own systems — a script, a CI job, an operations portal, a scheduler — over HTTP.
| You want to | You call |
|---|---|
| Ask a mission a question and get its answer | POST /missions/{id}/chat, then GET …/chat/{conversationId}?wait=60 |
| See what is waiting for approval, and approve or reject it | GET /missions/{id}/decisions?status=pending, POST …/decisions/{decisionId}/decide |
| Feed events from your system into a mission | POST /missions/{id}/watchers/{watcherId}/events |
| Read and change a mission's records | GET/POST/PATCH/DELETE /missions/{id}/tables/{table}/rows… |
| See every mission, its status and health; pause or activate one | GET /missions, GET /missions/{id}/health, PATCH /missions/{id} |
| Change settings, connections, people, schedules, agents | /missions/{id}/settings, /connections, /roles, /schedules, /agents |
| Install a mission from a bundle, upgrade it, export it | POST /missions, POST /missions/{id}/upgrade, GET /missions/{id}/export |
| Change how a mission works: watchers, playbooks, rules, pages | /missions/{id}/watchers, /playbooks, /rules, /pages, … |
Every path below is under https://<your-missions-host>/missions/api/v1.
The complete machine-readable description — every field of every request and
response — is served by your own installation at
GET /missions/api/v1/openapi.json (OpenAPI 3.1). Load it into Postman,
Swagger UI or a client generator.
Everything the API does is the same code the Missions screens run: the same approval limits, the same mission lock, the same audit trail. A change made through the API appears in the mission's history under the token's user.
1. Tokens
Send a token on every call:
Authorization: Bearer <token>
(X-Iris-Api-Key: <token> also works.) There are two kinds.
Personal token iris_pat_… |
Service token iris_svc_… |
|
|---|---|---|
| Acts as | You | A StudioX user you create for the integration |
| Made by | Anyone, for themselves | A business administrator |
| Made in | Missions → Settings → API Access | Missions → Settings → Service tokens |
| Chat, building agents, MCP tools | Work while your Studio sign-in is recent | Always work — nobody needs to be signed in |
| Business | The one you were in when you made it | The service user's business |
| Use it for | Your own scripts, trying things out | Systems that run on their own |
A token can be limited to what it may do (scopes) and given an expiry. A token with no scopes may do everything its user may do.
| Scope | Allows |
|---|---|
runtime |
Chat, approvals, watchers runs and events, reading mission data, files, health and usage |
manage |
Status, settings, connections, people, schedules, agents; raising a decision for a person; reading how a mission is built |
build |
Create, upgrade, clone, export missions; create and change watchers, playbooks, rules, procedures, pages, gates, ontologies, checkpoints |
data:write |
Add, change and delete mission rows |
A service token, or any token limited by scopes, works on /missions/api/v1
only. A service token sees its own business's missions and no others.
Tokens are made and revoked only in the Missions screens, never through the API, so a token that leaks cannot make more. A revoked token stops working within a minute. Only a hash of each token is stored; it is shown once, when it is made.
Setting up a service token
- In Studio, create a user for the integration — for example
missions-api@<your-company>— in your business, with the role the integration needs (a business administrator role if it will change settings, people or mission data). - Sign in to Studio as that user and generate its API key: Studio → Settings → Profile → API Key → Generate. Copy it and sign out. (A business administrator can always generate one; other roles only where the business allows API access.)
- Sign in as a business administrator, open Missions → Settings → Service tokens, paste the service user's API key, choose what the token may do and, optionally, when it expires.
- Tick Run this business's background work as this user if you want watchers, webhooks, playbooks and approvals from email to run as the service user instead of as whoever last opened each mission. One token per business can hold this.
- Copy the
iris_svc_…token. It is shown once.
If the service user's API key is generated again in Studio, the old key stops working and so does the service token: make a new service token with the new key.
Check a token with:
export MISSIONS="https://<your-missions-host>/missions/api/v1"
export TOKEN="iris_svc_…"
curl -s -H "Authorization: Bearer $TOKEN" "$MISSIONS/me"
{ "email": "missions-api@example.com", "business": "…", "tokenKind": "service",
"scopes": ["runtime", "manage"], "role": "business-admin",
"studioxCredential": "ok", "missionCount": 12 }
studioxCredential says whether calls that reach the agents, StudioX or MCP
tools will work now: ok, missing or expired. For a service token it is
always ok while the service user's key is current.
2. How the API behaves
JSON in, JSON out. File uploads are multipart/form-data; exports are
application/zip.
Errors are always { "error": "<what went wrong>", "code": "<code>" },
sometimes with more detail (missingSettings, scope, reason, details).
| Status | code |
Means |
|---|---|---|
| 400 | invalid_request |
The request is wrong; error says how |
| 401 | — | No token, or it is unknown, revoked or expired |
| 403 | scope_missing |
The token is not allowed this; scope names what it needs |
| 403 | forbidden |
Your role does not allow this (e.g. business administrators only) |
| 403 | locked |
The mission is locked against changes of this kind |
| 404 | not_found |
It does not exist, or you cannot see it |
| 409 | conflict |
Someone changed it since you read it (see below), or it is in a state that refuses this |
| 409 | studiox_session_stale |
A personal token whose Studio sign-in has lapsed; sign in once, or use a service token |
| 429 | rate_limited |
Over 600 requests a minute for this token; wait Retry-After seconds |
| 502 | upstream_error |
A connected system (an MCP server) refused or failed |
Lists take page (from 1) and pageSize, and answer with the page and a
total.
Status of a mission is draft, active or paused.
Who may change a mission. Running a mission — chat, approvals, events,
its records — needs only that you can see it. Changing how it is set up or
built (anything needing manage or build) also needs edit rights on it: you
created it, it is shared with you for editing, or you are a business
administrator. A service user is usually made a business administrator for
this reason.
Long operations run as jobs. Creating, upgrading or rebuilding a mission
answers 202 at once:
{ "jobId": "5e0c…", "status": "queued", "kind": "create-mission", "poll": "/missions/api/v1/jobs/5e0c…" }
Read GET /jobs/{jobId} until status is succeeded or failed. steps
shows progress; result is the outcome; error says why it failed.
Changing what someone else may be changing. Read the object, keep its
updatedAt, and send it back as expectedUpdatedAt with your change. If
someone changed it in between you get 409 conflict instead of overwriting
their work.
Audit. Every change, every chat and every MCP tool run through the API is recorded in the mission's audit log with the token's user and kind.
3. Running a mission
Ask a question
MID="<mission id>"
CID=$(curl -s -X POST -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
-d '{"message":"Which cases breached their SLA this week?"}' \
"$MISSIONS/missions/$MID/chat" | jq -r .conversationId)
curl -s -H "Authorization: Bearer $TOKEN" "$MISSIONS/missions/$MID/chat/$CID?wait=60" | jq -r .reply
?wait=60 holds the request until the answer is ready (up to 60 s); without
it the call answers at once with status: "running" and you call again. Send
conversationId with the next message to continue the same conversation.
GET …/chat/{conversationId}/trace shows how the answer was reached: which
agents and tools were used and for how long. When a conversation has been
handed to a person, the reply says status: "with-a-person".
Approvals
# what is waiting
curl -s -H "Authorization: Bearer $TOKEN" "$MISSIONS/missions/$MID/decisions?status=pending&pageSize=50"
# one decision, and who may approve it
curl -s -H "Authorization: Bearer $TOKEN" "$MISSIONS/missions/$MID/decisions/$DID"
# approve it
curl -s -X POST -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
-d '{"action":"approve","notes":"Checked against the contract"}' \
"$MISSIONS/missions/$MID/decisions/$DID/decide"
The approval limits set on the mission's roles apply exactly as in the
approvals queue: above your limit you get 403 with the limit; a decision
that needs a second approver records yours and answers awaiting. A decision
with options takes {"optionId": "…"} instead of action. Your system can
also raise a decision for a person to make: POST /missions/{id}/decisions
with event, what and why (scope manage); the mission's reviewers are
emailed as usual.
Feed events in
A watcher set to be triggered by a webhook takes your rows:
curl -s -X POST -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
-d '[{"alarm_id":"A-1042","site":"ESP-17","severity":"major","raised_at":"2026-10-16T13:05:00Z"}]' \
"$MISSIONS/missions/$MID/watchers/$WID/events"
The body is an array of rows, an object holding them under the watcher's rows
key, or a single row. Up to 500 rows a call; sending the same row twice stores
it once. POST …/watchers/{watcherId}/run runs any watcher now;
GET …/watchers/{watcherId}/logs shows what its runs did.
A system that cannot hold a token can still push to the watcher's own webhook
URL; a business administrator finds those URLs at GET /missions/{id}/endpoints.
Mission records
curl -s -H "Authorization: Bearer $TOKEN" "$MISSIONS/missions/$MID/tables"
curl -s -H "Authorization: Bearer $TOKEN" \
"$MISSIONS/missions/$MID/tables/cases/rows?filter[status]=open&sort=opened_at&dir=desc&pageSize=100"
curl -s -X PATCH -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
-d '{"set":{"status":"closed"}}' "$MISSIONS/missions/$MID/tables/cases/rows/CASE-118"
Reading and writing records is for business administrators, as on the
Database page. Writing needs the data:write scope and follows the same rules
as the mission's own agents: only the mission's columns, never a
credential column, every change recorded in the mission's history. A mission
marked sensitive, or one whose design forbids exporting its data, does not
open its records to the API.
How records are connected
Once the mission's tables are linked to its ontology (the Ontologies page, or
PUT /missions/{id}/ontology/mappings), ask how your records are connected.
With a question, the AI plans the walk and code runs it; with a plan, nothing
but code runs:
curl -s -X POST -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
-d '{"question":"Which customers are affected if router PE-SEA-01 goes down?"}' \
"$MISSIONS/missions/$MID/ontology/traverse"
curl -s -X POST -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
-d '{"plan":{"start":{"table":"routers","key":"PE-SEA-01"},
"path":[{"via":"dependsOn","direction":"in","maxHops":3},{"via":"serves","direction":"out"}],
"target":"Customer"}}' \
"$MISSIONS/missions/$MID/ontology/traverse"
Each result carries its path (pe1.sea1 (PE-SEA-01) ⟵ depends on ⟵ … ⟶ serves ⟶ Northwind Traders (CUST-01)), and plan in the answer is the plan that ran, to
keep and replay. "out" follows the record's own link column; "in" finds the
records whose link names it. A walk reads only. Uploading an ontology under a
name the mission already has makes its next version: list them at
…/ontologies/{ontologyId}/versions, compare two at …/diff?from=1&to=3,
restore one with …/restore, download its file at …/source.
4. Seeing and changing missions
curl -s -H "Authorization: Bearer $TOKEN" "$MISSIONS/missions?include=readiness"
curl -s -H "Authorization: Bearer $TOKEN" "$MISSIONS/missions/$MID/health"
# pause, then activate
curl -s -X PATCH -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
-d '{"status":"paused"}' "$MISSIONS/missions/$MID"
curl -s -X PATCH -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
-d '{"status":"active"}' "$MISSIONS/missions/$MID"
Activation is refused while a required setting is empty; the answer lists them
in missingSettings. Set them with PUT /missions/{id}/settings:
curl -s -X PUT -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
-d '{"values":{"inbox":"noc-cases@example.com"},"secrets":{"servicenow":{"password":"…"}}}' \
"$MISSIONS/missions/$MID/settings"
Secrets can be written but are never returned. When the mission's review policy sends settings changes to review, the answer is the change request rather than the change.
5. Building and updating missions
Install a mission from a bundle
# what the bundle needs from you
curl -s -X POST -H "Authorization: Bearer $TOKEN" -F file=@network-desk.sxm "$MISSIONS/bundles/inspect"
# install it, connect it, set it up, and switch it on — one job
curl -s -X POST -H "Authorization: Bearer $TOKEN" -F file=@network-desk.sxm \
-F 'options={"name":"Network Desk",
"integrationBindings":{"servicenow":"https://mcp.example.com/servicenow/mcp"},
"settings":{"inbox":"noc-cases@example.com"},
"enableWatchers":true,"activate":true}' \
"$MISSIONS/missions"
# → 202 { "jobId": "…", "poll": "/missions/api/v1/jobs/…" }
curl -s -H "Authorization: Bearer $TOKEN" "$MISSIONS/jobs/$JOB"
The job runs what the install screens run: import, settings, connections,
building the agents, switching on watchers, activation. result.mission is
the new mission. If activation was refused for an empty setting, the job still
succeeds with the mission created, the activate step failed and the list in
result.missingSettings. A mission can also be created from a blueprint
({"blueprint": {…}}) or from the prebuilt gallery (GET /prebuilts, then
{"prebuiltId": "…"}).
Change a mission's design and upgrade it
curl -s -H "Authorization: Bearer $TOKEN" "$MISSIONS/missions/$MID/blueprint" > blueprint.json
# edit blueprint.json
curl -s -X POST -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
-d "{\"blueprint\": $(jq .blueprint blueprint.json), \"dryRun\": true}" "$MISSIONS/missions/$MID/upgrade"
curl -s -X POST -H "Authorization: Bearer $TOKEN" -H "Content-Type: application/json" \
-d "{\"blueprint\": $(jq .blueprint blueprint.json)}" "$MISSIONS/missions/$MID/upgrade"
dryRun shows what would change without changing anything. An upgrade only
adds: new tables, columns, settings, watchers, agents and knowledge; pages are
rebuilt. Data, decisions and setting values are never touched, and anything
the new design no longer has is reported in findings and left in place. New
watchers arrive switched off — switch them on with
PATCH /missions/{id}/watchers/{watcherId} {"isEnabled": true}.
GET /missions/{id}/export?includeData=true downloads the mission as a bundle
to install elsewhere; POST /missions/{id}/clone copies it here.
6. Reference
Generated from the server's own catalog by node scripts/v1-reference-md.mjs — 160 operations. Request and response bodies: GET /missions/api/v1/openapi.json.
Platform
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/version |
Version and build of this Missions server | any | |
GET |
/openapi.json |
This API as OpenAPI 3.1 | any | |
GET |
/me |
Who the token acts as, and what it may do | any |
Jobs
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/jobs |
Jobs you started, newest first | any | |
GET |
/jobs/{jobId} |
One job: status, steps, result or error | any |
Chat
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
POST |
/missions/{id}/chat |
Ask the mission a question | runtime |
StudioX credential |
GET |
/missions/{id}/chat/{conversationId} |
Read a conversation's reply | runtime |
|
GET |
/missions/{id}/chat/{conversationId}/trace |
How the mission reached its reply | runtime |
|
POST |
/missions/{id}/agents/{agentId}/chat |
Ask one agent directly | runtime |
StudioX credential |
Conversations
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/conversations |
The mission's conversations | runtime |
|
DELETE |
/missions/{id}/conversations/{conversationId} |
Delete one of your conversations | runtime |
Decisions
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/decisions |
Decisions waiting for a person, and decided ones | runtime |
|
POST |
/missions/{id}/decisions |
Ask a person to decide something | manage |
|
GET |
/missions/{id}/decisions/summary |
How many decisions wait for you, and the latest | runtime |
|
GET |
/missions/{id}/decisions/{decisionId} |
One decision, and who may approve it | runtime |
|
POST |
/missions/{id}/decisions/{decisionId}/decide |
Approve, reject or choose an option | runtime |
Watchers
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/observations |
Watchers with their last run, and recent agent runs | runtime |
|
GET |
/missions/{id}/watchers |
The mission's watchers | runtime |
|
POST |
/missions/{id}/watchers |
Add a watcher | build |
|
GET |
/missions/{id}/watchers/{watcherId} |
One watcher, its recent runs, and what a run reads | runtime |
|
PATCH |
/missions/{id}/watchers/{watcherId} |
Change a watcher, or turn it on or off | any | |
DELETE |
/missions/{id}/watchers/{watcherId} |
Remove a watcher | build |
|
POST |
/missions/{id}/watchers/{watcherId}/run |
Run a watcher now | runtime |
|
POST |
/missions/{id}/watchers/{watcherId}/reset |
Let a watcher look at everything again | runtime or data:write |
|
GET |
/missions/{id}/watchers/{watcherId}/logs |
A watcher's log | runtime |
|
POST |
/missions/{id}/watchers/{watcherId}/events |
Push rows into a webhook watcher | runtime |
|
GET |
/missions/{id}/activity |
Recent watcher runs and findings | runtime |
Mission data
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/tables |
The mission's tables and columns | runtime |
|
GET |
/missions/{id}/tables/{table}/rows |
Rows of a mission table | runtime |
|
POST |
/missions/{id}/tables/{table}/rows |
Add a row | data:write |
|
GET |
/missions/{id}/tables/{table}/rows/{key} |
One row by its key | runtime |
|
PATCH |
/missions/{id}/tables/{table}/rows/{key} |
Change a row | data:write |
|
DELETE |
/missions/{id}/tables/{table}/rows/{key} |
Delete a row | data:write |
Files
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
POST |
/missions/{id}/files |
Upload a file for a mission form | runtime |
file upload |
GET |
/missions/{id}/files/{fileId} |
Download a file attached in a mission form | runtime |
application/octet-stream |
Procedures
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/procedure-runs |
Runs of the mission's procedures | runtime |
|
GET |
/missions/{id}/procedures |
The mission's procedures, with what its agents can do | build or manage |
|
POST |
/missions/{id}/procedures |
Enable a library procedure here, or write one on this mission | build |
|
GET |
/missions/{id}/procedures/{procedureId} |
One procedure, with its checks and any template update | build or manage |
StudioX credential |
PATCH |
/missions/{id}/procedures/{procedureId} |
Save the mission's copy: steps, wiring, on/off, follow updates | build |
|
DELETE |
/missions/{id}/procedures/{procedureId} |
Remove a procedure from the mission | build |
|
POST |
/missions/{id}/procedures/{procedureId}/validate |
Check a procedure (and optionally a draft of it) | build |
StudioX credential |
Health and usage
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/health |
Is the mission working right now | runtime |
|
GET |
/missions/{id}/cost |
What the mission has cost | runtime |
|
GET |
/missions/{id}/usage |
The mission's token use, by part and by person | runtime |
|
GET |
/missions/{id}/audit |
The mission's audit log | runtime |
Missions
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions |
The missions of your business, and your own | any | |
GET |
/missions/{id} |
One mission: its agents, readiness, install progress, lock and clock | any | |
PATCH |
/missions/{id} |
Rename, describe, activate or pause a mission | manage |
|
DELETE |
/missions/{id} |
Delete a mission and what it made | manage |
StudioX credential |
GET |
/missions/{id}/readiness |
What blocks activation | manage or build |
|
GET |
/missions/{id}/provisioning |
Progress of the mission's background install | manage or build |
Settings
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/settings |
The mission's settings: definitions, values, integrations, readiness | manage or build |
business admin |
PUT |
/missions/{id}/settings |
Save settings and credentials, and optionally activate | manage |
|
GET |
/missions/{id}/config |
The reasoning core's overrides for this mission | manage or build |
|
PUT |
/missions/{id}/config |
Set or reset reasoning core overrides | manage |
|
GET |
/missions/{id}/autonomy |
When each decision type may act without an approval | manage or build |
|
PUT |
/missions/{id}/autonomy |
Save one decision type's scenarios | manage |
business admin |
GET |
/missions/{id}/review-policy |
Which changes need a change request | manage or build |
|
PATCH |
/missions/{id}/review-policy |
Turn change requests on or off per surface | manage |
business admin |
Connections
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/connections |
The servers the mission's connections point at, each checked live | manage or build |
StudioX credential |
POST |
/missions/{id}/connections |
Check a server before pointing a connection at it | manage |
StudioX credential |
PUT |
/missions/{id}/connections |
Point a connection at a server | manage |
StudioX credential |
PATCH |
/missions/{id}/connections |
Mark a connection "not connected yet" | manage |
|
GET |
/missions/{id}/datasources |
The mission's data sources, credentials masked | manage or build |
|
PUT |
/missions/{id}/datasources/{kind} |
Create or change one data source | manage |
|
DELETE |
/missions/{id}/datasources/{kind} |
Remove one data source | manage |
|
POST |
/missions/{id}/datasources/{kind}/rotate-token |
New ingest token for the event stream (shown once) | manage |
|
POST |
/missions/{id}/datasources/{kind}/test |
Test a data source the way a watcher calls it | manage |
People
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/roles |
The mission's roles, what each may do, and who holds them | manage or build |
|
POST |
/missions/{id}/roles |
Give a person a role on the mission | manage |
business admin |
GET |
/missions/{id}/reviewers |
The people on the mission's roster | manage or build |
|
POST |
/missions/{id}/reviewers |
Add or update a person on the roster | manage |
business admin |
DELETE |
/missions/{id}/reviewers/{reviewerId} |
Take a person off the roster | manage |
business admin |
GET |
/missions/{id}/sharing |
Who can open the mission | manage or build |
|
POST |
/missions/{id}/sharing |
Change visibility, share or unshare | manage |
Schedules
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/schedules |
The mission's report schedules | manage or build |
|
POST |
/missions/{id}/schedules |
Create a report schedule | manage |
|
PATCH |
/missions/{id}/schedules/{scheduleId} |
Pause, resume or edit a schedule | manage |
|
DELETE |
/missions/{id}/schedules/{scheduleId} |
Delete a schedule | manage |
|
POST |
/missions/{id}/schedules/{scheduleId}/send |
Send a schedule once, now, to yourself | manage |
StudioX credential |
Agents
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/agents |
The mission's agents | any | |
POST |
/missions/{id}/agents |
Add an agent | manage |
|
GET |
/missions/{id}/agents/{agentId} |
One agent | any | |
PATCH |
/missions/{id}/agents/{agentId} |
Edit an agent | manage |
|
DELETE |
/missions/{id}/agents/{agentId} |
Remove an agent | build |
StudioX credential |
GET |
/missions/{id}/agents/{agentId}/settings |
The agent's model, limits, MCP tool timeout and specialization | manage or build |
StudioX credential |
PATCH |
/missions/{id}/agents/{agentId}/settings |
Change the agent's model, limits, MCP tool timeout or specialization | manage |
StudioX credential |
GET |
/missions/{id}/agents/{agentId}/tables |
What the agent may do with the mission's tables | manage or build |
|
PATCH |
/missions/{id}/agents/{agentId}/tables |
Let the agent read or change the mission's tables, or neither | manage |
StudioX credential |
GET |
/missions/{id}/agents/{agentId}/knowledge |
The documents on the agent's bot | manage or build |
StudioX credential |
POST |
/missions/{id}/agents/{agentId}/knowledge |
Add a plain-text document to the agent's bot | manage |
StudioX credential |
PATCH |
/missions/{id}/agents/{agentId}/knowledge/{sourceId} |
Switch a document on or off, or replace its text | manage |
StudioX credential |
DELETE |
/missions/{id}/agents/{agentId}/knowledge/{sourceId} |
Remove a document from the agent's bot | manage |
StudioX credential |
GET |
/missions/{id}/agents/{agentId}/mcp |
MCP servers you may use, each marked on or off for this agent | manage or build |
StudioX credential |
POST |
/missions/{id}/agents/{agentId}/mcp |
Switch an MCP server on or off for this agent | manage |
StudioX credential |
Keys
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
POST |
/missions/{id}/agents/{agentId}/webhook-secret |
New secret for the agent's webhook (shown once) | manage |
|
POST |
/missions/{id}/api-key |
New API key for the mission (shown once) | manage |
|
GET |
/missions/{id}/endpoints |
The URLs other systems push to | manage or build |
business admin |
Build
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/prebuilts |
The Prebuilt Missions gallery | any | |
GET |
/integrations |
Standard MCP integrations a blueprint can name, with their credential keys | build or manage |
|
POST |
/missions |
Create a mission from a bundle, a prebuilt or a blueprint | build |
StudioX credential, job (202), file upload |
POST |
/missions/{id}/upgrade |
Update a mission to a newer bundle or an edited blueprint | build |
StudioX credential, job (202), file upload |
POST |
/missions/{id}/finalize |
Build or repair the mission's bots and wiring | build |
StudioX credential, job (202) |
POST |
/missions/{id}/clone |
Copy a mission into a new one | build |
StudioX credential |
GET |
/missions/{id}/blueprint |
The mission's installed blueprint | build |
Bundles
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
POST |
/bundles/inspect |
Read a bundle and list what it asks before install | build |
file upload |
GET |
/missions/{id}/export |
Download the mission as a .sxm | build |
StudioX credential, application/zip |
Playbooks
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/playbooks |
The mission's playbooks, with how many watchers use each | build or manage |
|
POST |
/missions/{id}/playbooks |
Create a playbook | build |
|
GET |
/missions/{id}/playbooks/{playbookId} |
One playbook | build or manage |
|
PATCH |
/missions/{id}/playbooks/{playbookId} |
Change a playbook's name, description or content | build |
|
DELETE |
/missions/{id}/playbooks/{playbookId} |
Delete a playbook | build |
Rules
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/rules |
The mission's routing rules (which agent a message goes to) | build or manage |
|
POST |
/missions/{id}/rules |
Add a routing rule | build |
|
PATCH |
/missions/{id}/rules/{ruleId} |
Change a routing rule, or turn it on or off | build |
|
DELETE |
/missions/{id}/rules/{ruleId} |
Delete a routing rule | build |
|
GET |
/missions/{id}/formula-rules |
Every formula rule across the mission's watchers | build or manage |
|
PUT |
/missions/{id}/formula-rules |
Change one formula rule's expression | build |
business admin |
POST |
/missions/{id}/formula-rules/test |
Try a formula against the current rows without saving it | build |
Pages
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/pages |
The mission's pages | build or manage |
|
POST |
/missions/{id}/pages |
Add a page | build |
|
GET |
/missions/{id}/pages/{pageId} |
One page and its spec | build or manage |
|
PUT |
/missions/{id}/pages/{pageId} |
Replace a page's spec | build |
|
DELETE |
/missions/{id}/pages/{pageId} |
Delete a page | build |
Gates
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/gates |
The mission's gates | build or manage |
|
POST |
/missions/{id}/gates |
Add a gate | build |
|
GET |
/missions/{id}/gates/{gateId} |
One gate and every target it has judged | build or manage |
|
PATCH |
/missions/{id}/gates/{gateId} |
Change a gate | build |
|
DELETE |
/missions/{id}/gates/{gateId} |
Delete a gate | build |
|
POST |
/missions/{id}/gates/{gateId}/trigger |
Run a gate against a target | runtime |
|
POST |
/missions/{id}/gates/{gateId}/override |
Override a gate for a target | manage or build |
Ontologies
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/ontologies |
The mission's ontologies | build or manage |
|
POST |
/missions/{id}/ontologies |
Upload an ontology | build |
file upload |
GET |
/missions/{id}/ontologies/{ontologyId} |
One ontology with 20 sample triples | build or manage |
|
DELETE |
/missions/{id}/ontologies/{ontologyId} |
Delete an ontology and its triples | build |
|
POST |
/missions/{id}/ontologies/{ontologyId}/query |
Ask an ontology one question | runtime |
|
GET |
/missions/{id}/ontologies/{ontologyId}/versions |
Every kept version of an ontology | build or manage |
|
GET |
/missions/{id}/ontologies/{ontologyId}/diff |
What changed between two versions | build or manage |
|
POST |
/missions/{id}/ontologies/{ontologyId}/restore |
Restore an earlier version | build |
|
GET |
/missions/{id}/ontologies/{ontologyId}/source |
Download the file as uploaded | build or manage |
text/plain |
GET |
/missions/{id}/ontology/mappings |
The mission's tables linked to its model | build or manage |
|
PUT |
/missions/{id}/ontology/mappings |
Link the mission's tables to its model | build |
|
POST |
/missions/{id}/ontology/traverse |
How things are connected | runtime |
Checkpoints
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/checkpoints |
The mission's checkpoints, newest first | build or manage |
|
POST |
/missions/{id}/checkpoints |
Take a checkpoint now | build |
|
GET |
/missions/{id}/checkpoints/compare |
What changed between two checkpoints, or since one | build or manage |
|
POST |
/missions/{id}/checkpoints/{seq}/restore |
Restore the configuration to a checkpoint | build |
business admin |
MCP servers
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/mcp/servers |
MCP servers connected to your business | runtime |
StudioX credential |
GET |
/mcp/servers/{id}/tools |
Tools of one MCP server | runtime |
StudioX credential |
POST |
/mcp/servers/{id}/tools/{tool} |
Run one tool on an MCP server | runtime |
StudioX credential |
Case Desk
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
GET |
/missions/{id}/case-desk/tools |
The desk's tools and their arguments | runtime |
|
POST |
/missions/{id}/case-desk/tools/{tool} |
Run one of the desk's tools as you | runtime |
|
GET |
/missions/{id}/case-desk/cases |
Cases you may see, newest first | runtime |
|
POST |
/missions/{id}/case-desk/cases |
File a case | runtime |
|
GET |
/missions/{id}/case-desk/cases/{caseNo} |
One case as you may see it | runtime |
|
POST |
/missions/{id}/case-desk/cases/{caseNo}/messages |
Add a note, send a reply, or post the requester's message | runtime |
|
POST |
/missions/{id}/case-desk/cases/{caseNo}/status |
Move a case to another status | runtime |
|
POST |
/missions/{id}/case-desk/cases/{caseNo}/documents |
Attach a file or a link to a case | runtime |
Test Impact Scout
| Method | Path | What it does | Scope | Needs |
|---|---|---|---|---|
POST |
/missions/{id}/test-impact-scout/plan |
The tests to run for a batch of PRs | runtime |