Actions

Mikting actions use the regular script fields supported by ItsMyBot, including triggers, conditions, target, delay, chance, cooldown, and every. Most accept args.follow-up-actions. Follow-ups run after success or failure. To handle an error, branch on an output variable or use a condition. Actions do not accept not-met-actions; that field belongs to conditions. String inputs support [[variables]], %placeholders%, and {{secret.NAME}}. storeAs creates namespaced outputs such as [[probe_ok]].

HTTP requests

`miktingHttp`

Send a request to any HTTP endpoint. A request body is optional. If included, choose one of json, form, or text.

- id: miktingHttp
  args:
    method: POST            # GET|POST|PUT|PATCH|DELETE|HEAD (default GET)
    url: 'https://api.example.com/v1/check'
    headers: { Authorization: 'Bearer {{secret.API_TOKEN}}' }
    query: { verbose: 'true' }
    json: { text: '%message_content%' }
    # form: { q: 'hello' }  # url-encoded alternative
    # text: 'raw body'      # plain-text alternative
    # auth: { type: bearer, tokenSecret: API_TOKEN }  # instead of headers
    # auth: { type: basic, userSecret: API_USER, passSecret: API_PASS }
    timeoutMs: 10000
    retries: 1
    storeAs: mod
    follow-up-actions:
      - id: miktingBranch
        args:
          when: '[[mod_ok]] == false'
          then: [{ id: miktingReply, args: { content: 'API error: [[mikting_error]]' } }]

The action sets mikting_status, mikting_ok, mikting_body, and, when the response is JSON, mikting_json plus flattened JSON paths. Response headers are available as mikting_headers_<name>. With storeAs: probe, the same outputs use the probe_ prefix. For example, a JSON field at results.0.flagged becomes [[probe_json_results_0_flagged]]. When using auth.tokenSecret, numbered keys rotate automatically. The chosen slot name appears in mikting_key_name and probe_key_name. See Secrets.

Validate a response

responseCheck validates the response body against a known contract. Supported checks include OpenAI moderation, OpenAI and Groq policy responses, and Jev policy responses. A check adds valid and flagged outputs to its namespace; Jev also adds score. HTTP ok means only that the status is 2xx. Check both valid and flagged before taking action. OpenAI and Groq policy checks expect a completed Chat Completions response whose message contains JSON with a boolean flagged and a 1 to 500 character reason. Jev validates TypeSafe’s response shape, including its model, Noul answer, score, and token counters. policyThreshold accepts a number from 0 to 1 and defaults to 0.80. A score at or above the threshold sets flagged to true. This value is a model estimate, not a guarantee of accuracy. Extra fields are allowed; missing fields and invalid types fail validation. See the TypeSafe Noul contract.

`miktingHttpPaginate`

Follow a next URL or an RFC 5988 Link header and combine results:

- id: miktingHttpPaginate
  args:
    method: GET
    url: 'https://api.example.com/items'
    headers: { Authorization: 'Bearer {{secret.API_TOKEN}}' }
    nextPath: 'paging.next'   # dot path to next URL in each JSON body
    linkHeader: true          # or follow RFC5988 Link: <u>; rel="next"
    maxPages: 5
    storeAs: items            # -> [[items]] merged array + [[mikting_pages]]

Array response bodies are merged directly. For object bodies, the action merges the data array when present; otherwise it appends the object. It sets mikting_page_count and the named result array.

Variables

miktingSet { key, value } stores a value in the current run.
miktingGet { key, storeAs? } copies a value to a new name or the same key. A missing key sets mikting_error.
miktingAppend { key, value, sep? } appends to an array or joins text with sep (default ,). It creates a missing variable.
miktingClear { keys } removes one or more variables. keys can be a string or list.
miktingMath { expr, storeAs? } evaluates a numeric filtrex expression. It defaults the output name to result and does not use JavaScript eval.

- id: miktingMath
  args: { expr: '[[wins]] * 10 + 5', storeAs: xp }

Text and data

miktingString supports upper, lower, trim, length, slice, replace, split, join, and regex. slice uses from and to; replace uses a literal pattern and replacement; split and join use sep. regex uses a JavaScript regular expression and returns the match or capture groups, or null when there is no match. miktingJson parses, stringifies, or queries JSON. The query operation uses a dot path such as users.0.name. miktingArray supports len, first, last, reverse, unique, join, includes, and at. join uses sep, includes uses item, and at uses a zero-based index. A direct array variable input preserves objects; other string inputs are templated first. miktingDate supports now, format, parse, and add. now returns epoch milliseconds; format returns ISO text; parse returns milliseconds; add shifts a timestamp by an amount in ms, s, m, h, or d. miktingRandom produces an inclusive integer, UUID, or selected choice. miktingHash supports SHA-256 and MD5. miktingEncode converts between Base64, URL encoding, and hexadecimal.

Flow control

These actions run nested action lists where noted. They do not accept their own follow-up lists unless stated. miktingBranch { when, then?, else? } evaluates a templated boolean expression, then runs one of its action lists:

- id: miktingBranch
  args:
    when: '[[mod_json_results_0_flagged]] == true'
    then:
      - id: deleteMessage
      - id: miktingReply
        args: { content: 'Flagged and logged.' }

miktingForEach { list, itemVar, actions, maxIter? } runs actions for each array item and sets the item variable and its zero-based index. The default iteration limit is 100. miktingWhile { cond, actions, maxIter? } defaults to 25 iterations. miktingTry { try, catch?, storeErrorAs? } runs a catch list when the try list fails. miktingSleep { ms } pauses for up to 60 seconds. miktingStop ends the current chain. miktingCallFunction calls a function defined in a library. miktingInclude { path } runs an {actions:[...]} YAML file confined to the bot working directory. See Libraries.

Discord actions

miktingReply sends a message using standard message arguments, including content, embeds, components, and ephemeral. It replies to an interaction when available and otherwise sends to the channel. Follow-ups receive the sent message context. miktingDm { user? } sends a direct message to the context member by default. Closed DMs set mikting_error. miktingReact { value } reacts to the current message. miktingPoll { question, options, durationH? } creates a native poll with 2 to 10 options and a duration from 1 to 336 hours, defaulting to 24. miktingTimeout { durationSec?, user?, reason? } applies a timeout to the context member by default. Zero removes a timeout. The duration is in seconds. miktingModerate supports ban, unban, kick, addRole, removeRole, nickname, and slowmode; role and nickname operations accept their corresponding values, ban accepts deleteMessagesSec, and slowmode requires a text channel context. miktingWebhookSend { url, content, username?, storeAs? } sends JSON to a Discord webhook. Keep its URL in a secret. miktingPaginate { pages, timeoutSec? } sends the first page and lets the trigger author navigate the pages; the current index is stored in paginate_page. miktingWaitFor { type, timeoutSec?, fromUser?, storeAs? } waits for a message or reaction. Message results include content and author; reaction results include the emoji. A timeout sets mikting_error. Reactions require a message context.

Store actions

Use miktingStoreSet, miktingStoreGet, miktingStoreIncr, miktingStoreListOp, and miktingStoreRemove for persistent data. The store supports global, guild, user, channel, and message scopes. Setting the remove key to * clears the selected scope. See Key-value store. Documentation home · Source Markdown

Getting Started

Scripting

HTTP requests

`miktingHttp`

Validate a response

`miktingHttpPaginate`

Variables

Text and data

Flow control

Discord actions

Store actions

​HTTP requests

​miktingHttp

​Validate a response

​miktingHttpPaginate

​Variables

​Text and data

​Flow control

​Discord actions

​Store actions

HTTP requests

`miktingHttp`

Validate a response

`miktingHttpPaginate`

Variables

Text and data

Flow control

Discord actions

Store actions