Agent setup

Settings

The sandbox's own configuration, plus what it has saved you and which rules fired

On this page(18 sections)

Read the agent execution domain
Choose the agent execution domain
Read the AI literacy notice and your confirmation
Confirm your own AI literacy notice
List AI literacy confirmations
How this sandbox is configured
Change the sandbox settings
What the token-saving measures were worth
Read a built-in system prompt
When each rule last did something
What each repository asks to run on its own code
The state of this sandbox's field notes
Offer this sandbox a clock, if it has none
Which words the editor uses for you here
Choose which words the editor uses for you here
What you said about the getting-started checklist here
Put away or bring back the getting-started checklist
Switch a repository's own checks on or off

Read and write the settings that govern how agents behave here. The other three routes are read-only reports on their effects: what the token-saving measures were actually worth, the text behind a built-in prompt, and when each rule last did something.

18 calls. Pick one to open it, or use the list on the right.

GET/settings/agent-domainRead the agent execution domain

What you send

Nothing. Call it as it is.

What comes back

Field	Type
`agentDomain`	"root" \| "unprivileged"

Try itanswered in this tab

curl

curl "$SANDBOX/settings/agent-domain" \
  -H "x-intentic-control: $INTENTIC_TOKEN"

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.agentDomain();

POST/settings/agent-domainChoose the agent execution domain

What you send

Field	Type	Where
`agentDomain`	"root" \| "unprivileged"	body

What comes back

Field	Type
`ok`Always true	true

Try itanswered in this tab

curl

curl -X POST "$SANDBOX/settings/agent-domain" \
  -H "x-intentic-control: $INTENTIC_TOKEN" \
  -H "content-type: application/json" \
  -d '{"agentDomain":"root"}'

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.setAgentDomain({
  "agentDomain": "root"
});

GET/settings/ai-literacy/{locale}Read the AI literacy notice and your confirmation

What you send

Field	Type	Where
`locale`required	"en" \| "pl" \| "de" \| "fr" … (5)	address

What comes back

Field	Type
`show`	boolean
`text`	string
`version`	string
`confirmedAt`	string

Try itanswered in this tab

curl

curl "$SANDBOX/settings/ai-literacy/en" \
  -H "x-intentic-control: $INTENTIC_TOKEN"

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.aiLiteracy({
  "locale": "en"
});

POST/settings/ai-literacy/confirmConfirm your own AI literacy notice

What you send

Field	Type	Where
`version`required	string	body
`locale`required	"en" \| "pl" \| "de" \| "fr" … (5)	body

What comes back

Field	Type
`show`	boolean
`text`	string
`version`	string
`confirmedAt`	string

Try itanswered in this tab

curl

curl -X POST "$SANDBOX/settings/ai-literacy/confirm" \
  -H "x-intentic-control: $INTENTIC_TOKEN" \
  -H "content-type: application/json" \
  -d '{"version":"1.4.0","locale":"en"}'

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.confirmAiLiteracy({
  "version": "1.4.0",
  "locale": "en"
});

GET/settings/ai-literacy/confirmationsList AI literacy confirmations

What you send

Nothing. Call it as it is.

What comes back

Field	Type
`records`	object[]
`member`	string
`version`	string
`confirmedAt`	string

Try itanswered in this tab

curl

curl "$SANDBOX/settings/ai-literacy/confirmations" \
  -H "x-intentic-control: $INTENTIC_TOKEN"

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.aiLiteracyConfirmations();

GET/settingsHow this sandbox is configured

Every setting that governs how agents behave here, with the defaults filled in for anything nobody has chosen.

What you send

Nothing. Call it as it is.

What comes back

Field	Type
`timezone`Which clock this sandbox's schedules are…	"" \| string
`skills`Which built-in tools are switched on	string[]
`personaRouting`Whether a new chat is matched…	boolean
`chatPages`Let the agent answer with a…	boolean
`documentSearch`Keep the text of every document…	boolean
`sleepWhenUnused`Let the computer this sandbox runs…	boolean
`hashlineEdits`Have the agent edit files by…	boolean
`systemPromptMode`Which instructions the agent starts from:…	"intentic" \| "claude" \| "custom"
`systemPrompt`Your own instructions, used only when…	string
`iqSearch`Teach the agent how to use…	boolean
`iqSearchHoldout`What share of conversations to run…	number
`toolSearch`Load the schemas of connected tools…	boolean
`toolSearchHoldout`What share of conversations to run…	number
`stableSystemPrompt`Keep the instructions identical between turns…	boolean
`stableSystemPromptHoldout`What share of conversations to run…	number
`coldTrim`Resume large Claude Code conversations with…	boolean
`coldTrimHoldout`What share of conversations to resume…	number
`coldTrimMinTokens`How many measured context tokens a…	integer
`workspaceMap`Open every conversation with a map…	boolean
`workspaceMapHoldout`What share of conversations to open…	number
`fieldNotes`Open every turn with a brief…	boolean
`fieldNotesBudget`How much of that brief to…	integer
`fieldNotesHoldout`What share of conversations to run…	number
`outputCleaners`Which command outputs to trim before…	string
`outputHoldout`What share of commands to leave…	number
`modelSetups`The model setups jobs run on:…	object[]
`id`Which setup this is	string
`name`What it is called on the…	string
`pins`The models it runs on, tried…	object[]
`provider`Which provider serves this work	string
`model`Which of its models	string
`effort`How hard this model should think,…	string
`thinking`Whether this model reasons before it…	boolean
`fast`Ask for this model's work at…	boolean
`harness`Which agentic loop runs it	"native" \| "claude-code" \| "codex" \| "opencode" … (7)
`roleSetups`Which setup each job runs on:…	object
`autoModelGuidance`What you would tell somebody choosing…	string
`changelogRepos`Which repositories keep a changelog, and…	string[]
`agentRetentionDays`How many days a finished conversation…	number
`heldCheckoutDays`How many quiet days a held…	integer
`limitPolicy`What happens to a turn a…	"wait" \| "resend" \| "move"
`outagePolicy`What happens to a turn the…	"wait" \| "retry"
`turnCapMinutes`The longest one turn may run…	integer
`stopPolicy`What happens to a turn that…	"wait" \| "retry"
`memoryPolicy`What happens to a message the…	"wait" \| "resend"
`limitHandoff`How a turn a spent usage…	"suggested" \| "carry" \| "trim" \| "summary"
`limitMoveCarryUnder`For the suggested hand-off: carry the…	integer
`handoffSummaryOver`For the suggested hand-off: at or…	integer
`keepWarm`Keeping an idle conversation's prompt cache…	object
`auto`Keep a Claude conversation's prompt cache…	boolean
`hours`How long an idle conversation is…	number
`minTokens`Only conversations at least this large,…	integer
`reserve`How much of an account's usage…	integer
`autoRepair`Whether main's CI failing is repaired…	boolean
`followOrigin`Whether each workspace repo keeps up…	boolean
`offload`Which work runs on a runner…	object
`commands`	object
`sessions`	"suggest" \| "auto" \| "off"
`continueWhenNeedMet`Whether a conversation carries on by…	boolean
`autoResumeOnRestart`Whether a turn killed by the…	boolean
`adoptedChecks`Which repositories may run the checks…	object
`rules`Standing decisions about the sandbox's own…	object[]
`id`	string
`label`	string
`moment`	"file.edited" \| "turn.ending" \| "agent.finished" \| "agent.landed"
`when`	object
`repo`	string
`paths`	string[]
`outcome`	"clean" \| "error" \| "conflict" \| "checks-failed"[]
`sample`	number
`action`	object
`when kind is "command"`	shape
`command`	string
`timeoutMs`	number
`when kind is "verdict"`	shape
`verdict`	"allow" \| "hold"
`when kind is "builtin"`	shape
`name`	"verify-ui-edits" \| "version-landed"
`enabled`	boolean
`automationFailureLimit`How many failures in a row…	number
`voice`How the agent's replies are spoken…	object
`replies`When you talk to the agent…	boolean
`interrupt`Stop speaking the moment you start…	boolean
`speed`How fast it speaks, 1 being…	number
`voices`Which voice speaks each language, where…	object
`cloud`A speech provider of your own,…	object
`provider`Which provider speaks	"elevenlabs" \| "cartesia" \| "deepgram"
`secret`The stored secret holding its API…	string
`model`Which of its models speaks	string
`voice`Which of its voices, by the…	string
`wake`Talking to the agent from across…	object
`word`The name you call the agent…	string
`retention`How long this sandbox keeps each…	object
`transcriptDays`How many days a conversation's transcript…	integer
`restorePointKeepDays`How many days the workspace's restore…	integer
`privacyVaultDays`How many days the privacy shield…	integer
`cacheDays`How many days converted copies of…	integer
`rawOutputDays`How many days the full, untrimmed…	integer
`audit`The audit trail: who changed a…	object
`retentionDays`How many days audit records are…	integer
`recordReads`Also record who opened which conversation,…	boolean
`forwardUrl`An HTTPS address each audit record…	"" \| string
`forwardSecret`The name of a stored secret…	string
`productAnalytics`Let Intentic's product analytics (usage events…	boolean
`aiNotice`A notice members confirm once, telling…	object
`show`Show members a short notice the…	boolean
`text`Your own wording for that notice,…	string
`admission`Whether work started from outside may…	object
`schedule`	"allow" \| "hold" \| "deny"
`event`	"allow" \| "hold" \| "deny"
`listener`	"allow" \| "hold" \| "deny"
`webchat`	"allow" \| "hold" \| "deny"
`issues`	"allow" \| "hold" \| "deny"
`workspace`	"allow" \| "hold" \| "deny"
`workflow`	"allow" \| "deny"
`actionRules`What an agent may do out…	object
`commandJudge`Whether a model reads your safety…	"off" \| "watch" \| "on"
`projectInstalls`What happens when an agent installs…	"automatic" \| "ask" \| "never"
`subagentsAtOnce`How many subagents may work at…	number
`subagentsPerTurn`How many a single turn may…	number
`subagentDepth`How many levels deep the delegation…	number
`childRerunHorizonHours`How long a subagent refused for…	integer
`subagentBudget`Start a subagent only where an…	boolean
`subagentReservePercent`How much of each allowance window,…	integer
`automationThreadsAtOnce`How many threads of one team-chat…	number

Try itanswered in this tab

curl

curl "$SANDBOX/settings" \
  -H "x-intentic-control: $INTENTIC_TOKEN"

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.get();

POST/settingsChange the sandbox settings

Writes the settings whole, so send the complete object rather than the fields you changed.

What you send

Field	Type	Where
`timezone`Which clock this sandbox's schedules are…	"" \| string	body
`skills`Which built-in tools are switched on	string[]	body
`personaRouting`Whether a new chat is matched…	boolean	body
`chatPages`Let the agent answer with a…	boolean	body
`documentSearch`Keep the text of every document…	boolean	body
`sleepWhenUnused`Let the computer this sandbox runs…	boolean	body
`hashlineEdits`Have the agent edit files by…	boolean	body
`systemPromptMode`Which instructions the agent starts from:…	"intentic" \| "claude" \| "custom"	body
`systemPrompt`Your own instructions, used only when…	string	body
`iqSearch`Teach the agent how to use…	boolean	body
`iqSearchHoldout`What share of conversations to run…	number	body
`toolSearch`Load the schemas of connected tools…	boolean	body
`toolSearchHoldout`What share of conversations to run…	number	body
`stableSystemPrompt`Keep the instructions identical between turns…	boolean	body
`stableSystemPromptHoldout`What share of conversations to run…	number	body
`coldTrim`Resume large Claude Code conversations with…	boolean	body
`coldTrimHoldout`What share of conversations to resume…	number	body
`coldTrimMinTokens`How many measured context tokens a…	integer	body
`workspaceMap`Open every conversation with a map…	boolean	body
`workspaceMapHoldout`What share of conversations to open…	number	body
`fieldNotes`Open every turn with a brief…	boolean	body
`fieldNotesBudget`How much of that brief to…	integer	body
`fieldNotesHoldout`What share of conversations to run…	number	body
`outputCleaners`Which command outputs to trim before…	string	body
`outputHoldout`What share of commands to leave…	number	body
`modelSetups`The model setups jobs run on:…	object[]	body
`id`requiredWhich setup this is	string	body
`name`What it is called on the…	string	body
`pins`The models it runs on, tried…	object[]	body
`provider`requiredWhich provider serves this work	string	body
`model`requiredWhich of its models	string	body
`effort`How hard this model should think,…	string	body
`thinking`Whether this model reasons before it…	boolean	body
`fast`Ask for this model's work at…	boolean	body
`harness`Which agentic loop runs it	"native" \| "claude-code" \| "codex" \| "opencode" … (7)	body
`roleSetups`Which setup each job runs on:…	object	body
`autoModelGuidance`What you would tell somebody choosing…	string	body
`changelogRepos`Which repositories keep a changelog, and…	string[]	body
`agentRetentionDays`How many days a finished conversation…	number	body
`heldCheckoutDays`How many quiet days a held…	integer	body
`limitPolicy`What happens to a turn a…	"wait" \| "resend" \| "move"	body
`outagePolicy`What happens to a turn the…	"wait" \| "retry"	body
`turnCapMinutes`The longest one turn may run…	integer	body
`stopPolicy`What happens to a turn that…	"wait" \| "retry"	body
`memoryPolicy`What happens to a message the…	"wait" \| "resend"	body
`limitHandoff`How a turn a spent usage…	"suggested" \| "carry" \| "trim" \| "summary"	body
`limitMoveCarryUnder`For the suggested hand-off: carry the…	integer	body
`handoffSummaryOver`For the suggested hand-off: at or…	integer	body
`keepWarm`Keeping an idle conversation's prompt cache…	object	body
`auto`Keep a Claude conversation's prompt cache…	boolean	body
`hours`How long an idle conversation is…	number	body
`minTokens`Only conversations at least this large,…	integer	body
`reserve`How much of an account's usage…	integer	body
`autoRepair`Whether main's CI failing is repaired…	boolean	body
`followOrigin`Whether each workspace repo keeps up…	boolean	body
`offload`Which work runs on a runner…	object	body
`commands`	object	body
`sessions`	"suggest" \| "auto" \| "off"	body
`continueWhenNeedMet`Whether a conversation carries on by…	boolean	body
`autoResumeOnRestart`Whether a turn killed by the…	boolean	body
`adoptedChecks`Which repositories may run the checks…	object	body
`rules`Standing decisions about the sandbox's own…	object[]	body
`id`required	string	body
`label`required	string	body
`moment`required	"file.edited" \| "turn.ending" \| "agent.finished" \| "agent.landed"	body
`when`	object	body
`repo`	string	body
`paths`	string[]	body
`outcome`	"clean" \| "error" \| "conflict" \| "checks-failed"[]	body
`sample`	number	body
`action`required	object	body
`when kind is "command"`	shape	body
`command`required	string	body
`timeoutMs`	number	body
`when kind is "verdict"`	shape	body
`verdict`required	"allow" \| "hold"	body
`when kind is "builtin"`	shape	body
`name`required	"verify-ui-edits" \| "version-landed"	body
`enabled`	boolean	body
`automationFailureLimit`How many failures in a row…	number	body
`voice`How the agent's replies are spoken…	object	body
`replies`When you talk to the agent…	boolean	body
`interrupt`Stop speaking the moment you start…	boolean	body
`speed`How fast it speaks, 1 being…	number	body
`voices`Which voice speaks each language, where…	object	body
`cloud`A speech provider of your own,…	object	body
`provider`requiredWhich provider speaks	"elevenlabs" \| "cartesia" \| "deepgram"	body
`secret`requiredThe stored secret holding its API…	string	body
`model`Which of its models speaks	string	body
`voice`Which of its voices, by the…	string	body
`wake`Talking to the agent from across…	object	body
`word`The name you call the agent…	string	body
`retention`How long this sandbox keeps each…	object	body
`transcriptDays`How many days a conversation's transcript…	integer	body
`restorePointKeepDays`How many days the workspace's restore…	integer	body
`privacyVaultDays`How many days the privacy shield…	integer	body
`cacheDays`How many days converted copies of…	integer	body
`rawOutputDays`How many days the full, untrimmed…	integer	body
`audit`The audit trail: who changed a…	object	body
`retentionDays`How many days audit records are…	integer	body
`recordReads`Also record who opened which conversation,…	boolean	body
`forwardUrl`An HTTPS address each audit record…	"" \| string	body
`forwardSecret`The name of a stored secret…	string	body
`productAnalytics`Let Intentic's product analytics (usage events…	boolean	body
`aiNotice`A notice members confirm once, telling…	object	body
`show`Show members a short notice the…	boolean	body
`text`Your own wording for that notice,…	string	body
`admission`Whether work started from outside may…	object	body
`schedule`	"allow" \| "hold" \| "deny"	body
`event`	"allow" \| "hold" \| "deny"	body
`listener`	"allow" \| "hold" \| "deny"	body
`webchat`	"allow" \| "hold" \| "deny"	body
`issues`	"allow" \| "hold" \| "deny"	body
`workspace`	"allow" \| "hold" \| "deny"	body
`workflow`	"allow" \| "deny"	body
`actionRules`What an agent may do out…	object	body
`commandJudge`Whether a model reads your safety…	"off" \| "watch" \| "on"	body
`projectInstalls`What happens when an agent installs…	"automatic" \| "ask" \| "never"	body
`subagentsAtOnce`How many subagents may work at…	number	body
`subagentsPerTurn`How many a single turn may…	number	body
`subagentDepth`How many levels deep the delegation…	number	body
`childRerunHorizonHours`How long a subagent refused for…	integer	body
`subagentBudget`Start a subagent only where an…	boolean	body
`subagentReservePercent`How much of each allowance window,…	integer	body
`automationThreadsAtOnce`How many threads of one team-chat…	number	body

What comes back

Field	Type
`ok`Always true	true

Try itanswered in this tab

curl

curl -X POST "$SANDBOX/settings" \
  -H "x-intentic-control: $INTENTIC_TOKEN" \
  -H "content-type: application/json" \
  -d '{"timezone":"","skills":["…","…"],"personaRouting":true,"chatPages":true,"documentSearch":true,"sleepWhenUnused":true,"hashlineEdits":false,"systemPromptMode":"intentic","systemPrompt":"","iqSearch":false,"iqSearchHoldout":0,"toolSearch":true,"toolSearchHoldout":0.5,"stableSystemPrompt":true,"stableSystemPromptHoldout":0.5,"coldTrim":true,"coldTrimHoldout":0.5,"coldTrimMinTokens":100000,"workspaceMap":false,"workspaceMapHoldout":0,"fieldNotes":false,"fieldNotesBudget":4000,"fieldNotesHoldout":0,"outputCleaners":"","outputHoldout":0,"modelSetups":[{"id":"a1b2c3d4","name":"nightly changelog","pins":[{"provider":"claude","model":"claude-sonnet-4-6","effort":"…","thinking":true,"fast":true,"harness":"native"},{"provider":"claude","model":"claude-haiku-4-6","effort":"…","thinking":true,"fast":true,"harness":"claude-code"}]},{"id":"e5f6a7b8","name":"release notes","pins":[{"provider":"claude","model":"claude-sonnet-4-6","effort":"…","thinking":true,"fast":true,"harness":"native"},{"provider":"claude","model":"claude-haiku-4-6","effort":"…","thinking":true,"fast":true,"harness":"claude-code"}]}],"roleSetups":{"src/app.ts":"…","README.md":"…"},"autoModelGuidance":"","changelogRepos":["…","…"],"agentRetentionDays":3,"heldCheckoutDays":14,"limitPolicy":"wait","outagePolicy":"wait","turnCapMinutes":30,"stopPolicy":"wait","memoryPolicy":"wait","limitHandoff":"suggested","limitMoveCarryUnder":150000,"handoffSummaryOver":700000,"keepWarm":{"auto":false,"hours":4,"minTokens":100000,"reserve":15},"autoRepair":true,"followOrigin":true,"offload":{"commands":{"src/app.ts":"…","README.md":"…"},"sessions":"suggest"},"continueWhenNeedMet":true,"autoResumeOnRestart":false,"adoptedChecks":{"src/app.ts":"…","README.md":"…"},"rules":[{"id":"a1b2c3d4","label":"Nightly changelog","moment":"file.edited","when":{"repo":"root","paths":["src/app.ts","README.md"],"outcome":["clean","error"],"sample":1},"action":{"kind":"command","command":"pnpm dev","timeoutMs":900000},"enabled":true},{"id":"e5f6a7b8","label":"Release notes","moment":"turn.ending","when":{"repo":"site","paths":["src/app.ts","README.md"],"outcome":["clean","error"],"sample":1},"action":{"kind":"command","command":"pnpm build","timeoutMs":900000},"enabled":true}],"automationFailureLimit":0,"voice":{"replies":true,"interrupt":false,"speed":1,"voices":{"src/app.ts":"…","README.md":"…"},"cloud":{"provider":"elevenlabs","secret":"…","model":"","voice":""},"wake":{"word":"Leon"}},"retention":{"transcriptDays":0,"restorePointKeepDays":30,"privacyVaultDays":0,"cacheDays":0,"rawOutputDays":0},"audit":{"retentionDays":365,"recordReads":false,"forwardUrl":"","forwardSecret":""},"productAnalytics":true,"aiNotice":{"show":false,"text":""},"admission":{"schedule":"allow","event":"allow","listener":"allow","webchat":"allow","issues":"allow","workspace":"allow","workflow":"allow"},"actionRules":{"src/app.ts":"allow","README.md":"hold"},"commandJudge":"off","projectInstalls":"automatic","subagentsAtOnce":20,"subagentsPerTurn":200,"subagentDepth":3,"childRerunHorizonHours":6,"subagentBudget":true,"subagentReservePercent":10,"automationThreadsAtOnce":3}'

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.set({
  "timezone": "",
  "skills": [
    "…",
    "…"
  ],
  "personaRouting": true,
  "chatPages": true,
  "documentSearch": true,
  "sleepWhenUnused": true,
  "hashlineEdits": false,
  "systemPromptMode": "intentic",
  "systemPrompt": "",
  "iqSearch": false,
  "iqSearchHoldout": 0,
  "toolSearch": true,
  "toolSearchHoldout": 0.5,
  "stableSystemPrompt": true,
  "stableSystemPromptHoldout": 0.5,
  "coldTrim": true,
  "coldTrimHoldout": 0.5,
  "coldTrimMinTokens": 100000,
  "workspaceMap": false,
  "workspaceMapHoldout": 0,
  "fieldNotes": false,
  "fieldNotesBudget": 4000,
  "fieldNotesHoldout": 0,
  "outputCleaners": "",
  "outputHoldout": 0,
  "modelSetups": [
    {
      "id": "a1b2c3d4",
      "name": "nightly changelog",
      "pins": [
        {
          "provider": "claude",
          "model": "claude-sonnet-4-6",
          "effort": "…",
          "thinking": true,
          "fast": true,
          "harness": "native"
        },
        {
          "provider": "claude",
          "model": "claude-haiku-4-6",
          "effort": "…",
          "thinking": true,
          "fast": true,
          "harness": "claude-code"
        }
      ]
    },
    {
      "id": "e5f6a7b8",
      "name": "release notes",
      "pins": [
        {
          "provider": "claude",
          "model": "claude-sonnet-4-6",
          "effort": "…",
          "thinking": true,
          "fast": true,
          "harness": "native"
        },
        {
          "provider": "claude",
          "model": "claude-haiku-4-6",
          "effort": "…",
          "thinking": true,
          "fast": true,
          "harness": "claude-code"
        }
      ]
    }
  ],
  "roleSetups": {
    "src/app.ts": "…",
    "README.md": "…"
  },
  "autoModelGuidance": "",
  "changelogRepos": [
    "…",
    "…"
  ],
  "agentRetentionDays": 3,
  "heldCheckoutDays": 14,
  "limitPolicy": "wait",
  "outagePolicy": "wait",
  "turnCapMinutes": 30,
  "stopPolicy": "wait",
  "memoryPolicy": "wait",
  "limitHandoff": "suggested",
  "limitMoveCarryUnder": 150000,
  "handoffSummaryOver": 700000,
  "keepWarm": {
    "auto": false,
    "hours": 4,
    "minTokens": 100000,
    "reserve": 15
  },
  "autoRepair": true,
  "followOrigin": true,
  "offload": {
    "commands": {
      "src/app.ts": "…",
      "README.md": "…"
    },
    "sessions": "suggest"
  },
  "continueWhenNeedMet": true,
  "autoResumeOnRestart": false,
  "adoptedChecks": {
    "src/app.ts": "…",
    "README.md": "…"
  },
  "rules": [
    {
      "id": "a1b2c3d4",
      "label": "Nightly changelog",
      "moment": "file.edited",
      "when": {
        "repo": "root",
        "paths": [
          "src/app.ts",
          "README.md"
        ],
        "outcome": [
          "clean",
          "error"
        ],
        "sample": 1
      },
      "action": {
        "kind": "command",
        "command": "pnpm dev",
        "timeoutMs": 900000
      },
      "enabled": true
    },
    {
      "id": "e5f6a7b8",
      "label": "Release notes",
      "moment": "turn.ending",
      "when": {
        "repo": "site",
        "paths": [
          "src/app.ts",
          "README.md"
        ],
        "outcome": [
          "clean",
          "error"
        ],
        "sample": 1
      },
      "action": {
        "kind": "command",
        "command": "pnpm build",
        "timeoutMs": 900000
      },
      "enabled": true
    }
  ],
  "automationFailureLimit": 0,
  "voice": {
    "replies": true,
    "interrupt": false,
    "speed": 1,
    "voices": {
      "src/app.ts": "…",
      "README.md": "…"
    },
    "cloud": {
      "provider": "elevenlabs",
      "secret": "…",
      "model": "",
      "voice": ""
    },
    "wake": {
      "word": "Leon"
    }
  },
  "retention": {
    "transcriptDays": 0,
    "restorePointKeepDays": 30,
    "privacyVaultDays": 0,
    "cacheDays": 0,
    "rawOutputDays": 0
  },
  "audit": {
    "retentionDays": 365,
    "recordReads": false,
    "forwardUrl": "",
    "forwardSecret": ""
  },
  "productAnalytics": true,
  "aiNotice": {
    "show": false,
    "text": ""
  },
  "admission": {
    "schedule": "allow",
    "event": "allow",
    "listener": "allow",
    "webchat": "allow",
    "issues": "allow",
    "workspace": "allow",
    "workflow": "allow"
  },
  "actionRules": {
    "src/app.ts": "allow",
    "README.md": "hold"
  },
  "commandJudge": "off",
  "projectInstalls": "automatic",
  "subagentsAtOnce": 20,
  "subagentsPerTurn": 200,
  "subagentDepth": 3,
  "childRerunHorizonHours": 6,
  "subagentBudget": true,
  "subagentReservePercent": 10,
  "automationThreadsAtOnce": 3
});

GET/settings/savingsWhat the token-saving measures were worth

Measured rather than estimated: what each mechanism actually saved over a range of days. The same day range the spending ledger takes, so one calendar filters both.

What you send

Field	Type	Where
`from`First day to include, as YYYY-MM-DD…	string	query
`to`Last day to include, as YYYY-MM-DD…	string	query

What comes back

Field	Type
`input`	object
`updatedAt`	number
`commands`	number
`rawTokens`	number
`emittedTokens`	number
`savedPct`	number
`perCleaner`	object[]
`id`	string
`commands`	number
`savedTokens`	number
`holdout`	object
`cleaned`	number
`heldOut`	number
`measuredSavedPct`	number
`gaps`	object[]
`command`	string
`commands`	number
`tokens`	number
`search`	object
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`turns`	number
`mean`	number
`off`	object
`turns`	number
`mean`	number
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`minTurns`	number
`sampleUnit`	"turns" \| "conversations" \| "opening turns"
`cohort`	string
`previous`	object
`from`	number
`to`	number
`cohort`	string
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`off`	object
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`uptake`	object
`on`	number
`off`	number
`turns`	number
`map`	object
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`turns`	number
`mean`	number
`off`	object
`turns`	number
`mean`	number
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`minTurns`	number
`sampleUnit`	"turns" \| "conversations" \| "opening turns"
`cohort`	string
`previous`	object
`from`	number
`to`	number
`cohort`	string
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`off`	object
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`uptake`	object
`on`	number
`off`	number
`turns`	number
`notes`	object
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`turns`	number
`mean`	number
`off`	object
`turns`	number
`mean`	number
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`minTurns`	number
`sampleUnit`	"turns" \| "conversations" \| "opening turns"
`cohort`	string
`previous`	object
`from`	number
`to`	number
`cohort`	string
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`off`	object
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`uptake`	object
`on`	number
`off`	number
`turns`	number
`toolSearch`	object
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`turns`	number
`mean`	number
`off`	object
`turns`	number
`mean`	number
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`minTurns`	number
`sampleUnit`	"turns" \| "conversations" \| "opening turns"
`cohort`	string
`previous`	object
`from`	number
`to`	number
`cohort`	string
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`off`	object
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`uptake`	object
`on`	number
`off`	number
`turns`	number
`stablePrompt`	object
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`turns`	number
`mean`	number
`off`	object
`turns`	number
`mean`	number
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`minTurns`	number
`sampleUnit`	"turns" \| "conversations" \| "opening turns"
`cohort`	string
`previous`	object
`from`	number
`to`	number
`cohort`	string
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`off`	object
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`uptake`	object
`on`	number
`off`	number
`turns`	number
`coldTrim`	object
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`turns`	number
`mean`	number
`off`	object
`turns`	number
`mean`	number
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`minTurns`	number
`sampleUnit`	"turns" \| "conversations" \| "opening turns"
`cohort`	string
`previous`	object
`from`	number
`to`	number
`cohort`	string
`metrics`	object[]
`metric`	"searchCalls" \| "openingSearches" \| "openedWithListing" \| "callsBeforeTarget" … (7)
`on`	object
`off`	object
`controlTurnsNeeded`	number
`marginPct`	number
`deltaPct`	number
`saved`	number
`uptake`	object
`on`	number
`off`	number
`turns`	number

Try itanswered in this tab

curl

curl "$SANDBOX/settings/savings" \
  -H "x-intentic-control: $INTENTIC_TOKEN"

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.savings();

GET/settings/system-prompt/{base}Read a built-in system prompt

The actual text behind one of the built-in modes, so a settings screen can show the prompt instead of asking anyone to trust a description of it, and so either can be forked into a custom one.

What you send

Field	Type	Where
`base`required	"intentic" \| "claude"	address

What comes back

Field	Type
`text`	string
`version`	string

Try itanswered in this tab

curl

curl "$SANDBOX/settings/system-prompt/intentic" \
  -H "x-intentic-control: $INTENTIC_TOKEN"

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.builtinPrompt({
  "base": "intentic"
});

GET/settings/rule-firingsWhen each rule last did something

A separate read rather than a field on the settings, because a rule firing is not somebody editing anything: folding it in would turn every firing into a settings write and put a self-changing value inside the object a screen edits.

What you send

Nothing. Call it as it is.

What comes back

A plain value rather than an object. The example below is the whole of it.

Try itanswered in this tab

curl

curl "$SANDBOX/settings/rule-firings" \
  -H "x-intentic-control: $INTENTIC_TOKEN"

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.firings();

GET/settings/repo-checksWhat each repository asks to run on its own code

Every repository that declares its own checks at `.intentic/checks.json`, what it declares, and whether you have switched it on. A repository declares what to run because the command belongs beside the scripts it names; nothing it declares runs until you say so.

What you send

Nothing. Call it as it is.

What comes back

Field	Type
`repos`Every repository that declares checks, in…	object[]
`repo`Which repository, by its workspace id…	string
`path`Where the declaration lives, relative to…	string
`checks`What it declares, in the order…	object[]
`when`When to run it: `edit` on…	"edit" \| "turn" \| "land"
`run`The command, run in this repository's…	string
`label`What to call it on screen	string
`timeoutMs`How long it may take before…	number
`paths`Only run it when the change…	string[]
`fired`When each declared check last reported…	number[]
`adopted`Whether these are running	boolean
`changed`Whether the declaration gained or rewrote…	boolean
`held`For a changed declaration, which of…	boolean[]
`error`Why the file could not be…	string

Try itanswered in this tab

curl

curl "$SANDBOX/settings/repo-checks" \
  -H "x-intentic-control: $INTENTIC_TOKEN"

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.repoChecks();

GET/settings/field-notesThe state of this sandbox's field notes

Whether `.intentic/config/field-notes.toon` exists, when it was last rewritten, how much of it the current budget reaches, and whether a monthly rewrite is scheduled.

What you send

Nothing. Call it as it is.

What comes back

Field	Type
`present`	boolean
`writtenAt`	number
`ranksSent`	number
`ranksTotal`	number
`chars`	number
`automation`	"missing" \| "enabled" \| "disabled"
`nextRunAt`	number
`unreadable`	string
`unlisted`	string[]

Try itanswered in this tab

curl

curl "$SANDBOX/settings/field-notes" \
  -H "x-intentic-control: $INTENTIC_TOKEN"

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.fieldNotes();

POST/settings/timezoneOffer this sandbox a clock, if it has none

Sets which timezone this sandbox's schedules are meant in, but only while it has none set. Already answered, the stored zone wins and comes back unchanged, so any number of browsers can offer theirs without fighting over it. To change a zone that is already set, write the settings.

What you send

Field	Type	Where
`timezone`requiredThe zone the offering machine is…	string	body

What comes back

Field	Type
`timezone`The zone this sandbox's schedules are…	string
`adopted`Whether this call is what set…	boolean

Try itanswered in this tab

curl

curl -X POST "$SANDBOX/settings/timezone" \
  -H "x-intentic-control: $INTENTIC_TOKEN" \
  -H "content-type: application/json" \
  -d '{"timezone":"…"}'

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.adoptTimezone({
  "timezone": "…"
});

GET/settings/audienceWhich words the editor uses for you here

Whether your editor speaks git's own words (developer) or plain ones (maker) on this sandbox, in every browser and on every device. Absent until you have chosen here.

What you send

Nothing. Call it as it is.

What comes back

Field	Type
`audience`developer for git's own words, maker…	"developer" \| "maker"

Try itanswered in this tab

curl

curl "$SANDBOX/settings/audience" \
  -H "x-intentic-control: $INTENTIC_TOKEN"

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.audience();

POST/settings/audienceChoose which words the editor uses for you here

Sets whether your editor speaks git's own words (developer) or plain ones (maker) on this sandbox, for every browser and device you open it on. Other members keep their own. With `offer` it is taken only while you have none kept, and the kept answer comes back unchanged.

What you send

Field	Type	Where
`audience`requireddeveloper for git's own words, maker…	"developer" \| "maker"	body
`offer`Take it only while this person…	boolean	body

What comes back

Field	Type
`audience`The words this sandbox's editor now…	"developer" \| "maker"
`adopted`Whether this call set it	boolean

Try itanswered in this tab

curl

curl -X POST "$SANDBOX/settings/audience" \
  -H "x-intentic-control: $INTENTIC_TOKEN" \
  -H "content-type: application/json" \
  -d '{"audience":"developer","offer":true}'

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.setAudience({
  "audience": "developer",
  "offer": true
});

GET/settings/getting-startedWhat you said about the getting-started checklist here

Whether you put the editor's getting-started checklist away on this sandbox, and which of its steps you passed over. Whether a step is done is not kept here: the editor reads that from the sandbox itself.

What you send

Nothing. Call it as it is.

What comes back

Field	Type
`hidden`Whether this person put the checklist…	boolean
`skipped`The checklist steps this person passed…	string[]

Try itanswered in this tab

curl

curl "$SANDBOX/settings/getting-started" \
  -H "x-intentic-control: $INTENTIC_TOKEN"

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.gettingStarted();

POST/settings/getting-startedPut away or bring back the getting-started checklist

Replaces what you said about the editor's getting-started checklist on this sandbox: whether it is put away, and which steps you passed over. Other members keep their own.

What you send

Field	Type	Where
`hidden`Whether this person put the checklist…	boolean	body
`skipped`The checklist steps this person passed…	string[]	body

What comes back

Field	Type
`hidden`Whether this person put the checklist…	boolean
`skipped`The checklist steps this person passed…	string[]

Try itanswered in this tab

curl

curl -X POST "$SANDBOX/settings/getting-started" \
  -H "x-intentic-control: $INTENTIC_TOKEN" \
  -H "content-type: application/json" \
  -d '{"hidden":true,"skipped":["…","…"]}'

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.setGettingStarted({
  "hidden": true,
  "skipped": [
    "…",
    "…"
  ]
});

POST/settings/repo-checks/adoptSwitch a repository's own checks on or off

Adopts exactly what that repository declares as it stands now. If the declaration changes afterwards it stops running until you adopt it again, so a command nobody has read cannot inherit the answer given to a different one.

What you send

Field	Type	Where
`repo`requiredWhich repository's declaration to switch	string	body
`on`requiredOn adopts what it declares as…	boolean	body

What comes back

Field	Type
`ok`Always true	true

Try itanswered in this tab

curl

curl -X POST "$SANDBOX/settings/repo-checks/adopt" \
  -H "x-intentic-control: $INTENTIC_TOKEN" \
  -H "content-type: application/json" \
  -d '{"repo":"root","on":true}'

TypeScript

import { sandbox } from "@intentic/sandbox-client";

const result = await sandbox.settings.adoptRepoChecks({
  "repo": "root",
  "on": true
});