NPCs

Contents

Creating new NPCs

Often you will want to create new NPCs to either add quests, flavor, or to introduce little known activities to the larger player base. New NPCs are written entirely in JSON, so they are one of the easier places to begin your contributions.

There are two parts to creating a new NPC, apart from any dialogue you may want to add.

NPC Class definition

First there is the npc_class which follows the following template. Format:

{
  "type": "npc_class",
  "id": "NC_EXAMPLE",                                                      // Mandatory, unique id that refers to this class.
  "name": { "str": "Example NPC" },                                        // Mandatory, display name for this class.
  "job_description": "I'm helping you learn the game.",                    // Mandatory
  "common": false,                                                         // Optional. Whether or not this class can appear via random generation.
  "sells_belongings": false,                                               // Optional. See [Shopkeeper NPC configuration](#shopkeeper-npc-configuration)
  "bonus_str": { "rng": [ -4, 0 ] },                                       // Optional. Modifies stat by the given value. This example shows a random distribution between -4 and 0.
  "bonus_dex": 100,                                                        // Optional. This example always adds exactly 100 to the stat.
  "bonus_int": { "one_in": 3 },                                            // Optional. This example adds 1 to the stat, but only about ~33.33% of the time (1 in 3).
  "bonus_per": { "sum": [ { "constant": 100 }, { "dice": [ 10, 10 ] } ] }, // Optional. This example adds 100 + 10d10 (10 dice each with 10 sides) to the stat.
  "bonus_aggression": { "rng": [ -4, 0 ] },                                // Optional. Modifies NPC's personality score like the above examples. Resulting value will be
                                                                           // clamped to within a range of -10, 10. (e.g. if aggression would be -12 it's instead set to -10)
  "bonus_bravery": 100,                                                    // Optional.
  "bonus_collector": { "one_in": 3 },                                      // Optional.
  "bonus_altruism": -10,                                                   // Optional.
  "skills": [
    {
      "skill": "ALL",                                                      // Optional. Applies bonuses/penalties to skills like the examples above. `ALL` is a special string which
                                                                           // applies to all skills not otherwise modified. See data/json/skills.json for a list of skill IDs.
      "level": { "mul": [ { "one_in": 3 }, { "sum": [ { "dice": [ 2, 2 ] }, { "constant": -2 }, { "one_in": 4 } ] } ] }
    }
  ],
  "worn_override": "NC_EXAMPLE_worn",                                      // Optional. Defines an [item group](/ITEM_SPAWN.html) that replaces all of their worn items.
  "carry_override": "NC_EXAMPLE_carried",                                  // Optional. Defines an item group that replaces their carried items (in pockets, etc). Items which cannot
                                                                           // be carried will overflow(fall to the ground) when the NPC is loaded.
  "weapon_override": "NC_EXAMPLE_weapon",                                  // Optional. Defines an item group that replaces their wielded weapon.
  "shopkeeper_item_group": [                                               // Optional. See [Shopkeeper NPC configuration](#shopkeeper-npc-configuration) below.
    { "group": "example_shopkeeper_itemgroup1" },
    { "group": "example_shopkeeper_itemgroup2", "trust": 10 },
    { "group": "example_shopkeeper_itemgroup3", "trust": 20, "rigid": true }
    { "group": "example_shopkeeper_itemgroup3", "trust": 40, "strict": true },
    {
      "group": "example_shopkeeper_itemgroup4",
      "condition": { "u_has_var": "VIP", "type": "general", "context": "examples", "value": "yes" }
    }
  ],
  "shopkeeper_consumption_rates": "basic_shop_rates",
  "shopkeeper_price_rules": [
    { "item": "scrap", "price": 10000 },
  ],
  "shopkeeper_blacklist": "test_blacklist",
  "restock_interval": "6 days",
  "traits": [ { "group": "BG_survival_story_EVACUEE" }, { "group": "NPC_starting_traits" }, { "group": "Appearance_demographics" } ]     // Optional
}

Shopkeeper NPC configuration

npc_class supports several properties for configuring the behavior of NPCs that behave as shopkeepers:

  • "sells_belongings": false means that this NPC’s worn or held items will strictly be excluded from their shopkeeper list; otherwise, they’ll be happy to sell things like their pants. It defaults to true if not specified.
  • "shopkeeper_item_group" is only needed if the planned NPC will be a shopkeeper with a revolving stock of items that change every three in-game days. All of the item overrides will ensure that any NPC of this class spawns with specific items.
  • "shopkeeper_consumption_rates" optional to define item consumption rates for this shopkeeper. Default is to consume all items before restocking
  • "shopkeeper_price_rules" optional to define personal price rules with the same format as faction price rules (see FACTIONS.md). These take priority over faction rules
  • "shopkeeper_blacklist" optional to define blacklists for this shopkeeper
  • "restock_interval": optional. Default is 6 days

Shopkeeper item groups

"shopkeeper_item_group" entries have the following fields:

  • "group" : Identifies an item group to include in the possible shop rotation
  • "trust" : (optional) If the faction’s trust with the player is below this value, items in this group will not be available for sale (Defaults to 0)
  • "condition" : (optional) Checked alongside trust with the avatar as alpha and the evaluating NPC as beta. See Player or NPC conditions.
  • "strict" : (optional) If true, items in this group will not be available for restocking unless the conditions are met. (Defaults to false)
  • "rigid" : (optional) By default, item groups will be continually iterated until they reach a certain value or size threshold for the NPC. Rigid groups are instead guaranteed to populate a single time if they can, and will not include duplicate reruns. (Defaults to false)
  • "refusal" : (optional) message to display in UIs (ex: trade UI) when conditions are not met. Defaults to "<npcname> does not trust you enough"

Shopkeeper consumption rates

Controls consumption of shopkeeper’s stock of items (simulates purchase by other people besides they player).

  "type": "shopkeeper_consumption_rates",
  "id": "basic_shop_rates",
  "default_rate": 5, // defined as units/day since last restock
  "junk_threshold": "10 cent", // items below this price will be consumed completely regardless of matches below
  "rates": [ // lower entries override higher ones
    { "item": "hammer", "rate": 1 },
    {
      "item": "hammer",
      "rate": 10,
      "condition": { "npc_has_var": "hammer_eater", "type": "bool", "context": "dinner", "value": "yes" }
    },
    { "category": "ammo", "rate": 10 },
    { "group": "EXODII_basic_trade", "rate": 100 }
    { "group": "EXODII_basic_trade", "category": "ammo", "rate": 200 }
  ]

condition is checked with avatar as alpha and npc as beta. See Player or NPC conditions.

Shopkeeper blacklists

Specifies blacklist of items that shopkeeper will not accept for trade. Format is similar to shopkeeper_consumption_rates.

  "type": "shopkeeper_blacklist",
  "id": "basic_blacklist",
  "entries": [
    {
      "item": "hammer",
      "condition": { "npc_has_var": "hammer_hater", "type": "bool", "context": "test", "value": "yes" },
      "message": "<npcname> hates this item"
    },
    { "category": "ammo" },
    { "group": "EXODII_basic_trade" }
  ]

Shop restocking

NPCs with at least one shopkeeper_item_group will (re)stock their shop in nearby loot zones (within PICKUP_RANGE = 6 tiles) owned by their faction and will ignore all other items. If there isn’t at least one LOOT_UNSORTED zone nearby, fallback zones will be automatically placed on all nearby, reachable, unsealed furniture with either the CONTAINER flag or a max volume higher than the floor. If there is no suitable furniture around, a 3x3 zone centered on the NPC will be created instead.

Before restocking, items owned by the NPC’s faction within these zones will be consumed according to shopkeeper_consumption_rates.

The shop restocks every restock_interval regardless of interactions with the avatar.

NOTE: do not place items within these loot zones in mapgen definitions as these will be consumed during the first restock. Add them to the item groups instead.

NPC instance definition

There is a second template required for a new NPC. It looks like this: Format:

{
  "type": "npc",
  "id": "examplicious",
  "//": "The luckiest NPC to never experience the Cataclysm.",
  "name_suffix": "examplar",
  "class": "NC_EXAMPLE",
  "attitude": 0,
  "mission": 7,
  "chat": "TALK_EXAMPLE",
  "faction": "no_faction",
  "death_eocs": [ "EOC_DEATH_NPC_TEST" ],
  "age": 30,
  "height": 180,
  "str": 7,
  "dex": 8,
  "int": 9,
  "per": 10,
  "personality": {
    "aggression": -1,
    "bravery":  2,
    "collector":  -3,
    "altruism":  4
  }
}

This is the JSON that creates the NPC ID that is used to spawn an NPC in “mapgen” (map generation). If optional fields are omitted, values are defined randomly.

field Description
name_unique Set name of NPC.
name_suffix Set name suffix of NPC.
attitude (mandatory) Based on the enum in npc.h. The important ones are 0=NPCATT_NULL, 1=NPCATT_TALK, 3=NPCATT_FOLLOW, 10=NPCATT_KILL, and 11=NPCATT_FLEE.
mission (mandatory) Based on the enum in npc.h. The important ones are 0=NPC_MISSION_NULL, 3=NPC_MISSION_SHOPKEEP, 7=NPC_MISSION_GUARD, and 8=NPC_MISSION_GUARD_PATROL.
chat (mandatory) Covered in the dialogue examples below.
faction Set faction NPC belongs to (see FACTIONS.md).
death_eocs String effect_on_condition ids and/or inline effect_on_conditions (see EFFECT_ON_CONDITION.md). When the npc dies all of these eocs are run with the victim as u and the killer as npc.
age Set age of NPC.
height Set height of NPC. (in cm)
str Set strength of NPC.
dex Set dexterity of NPC.
int Set intelligence of NPC.
per Set perception of NPC.
personality Set personality of NPC. For example, positive value of aggression means NPC is aggressive, negative value means NPC is defensive.

Writing dialogues

Dialogues work like state machines. They start with a certain topic (the NPC says something), the player character can then respond (choosing one of several responses), and that response sets the new talk topic. This goes on until the dialogue is finished, or the NPC turns hostile.

Note that it is perfectly fine to have a response that switches the topic back to itself.

NPC missions are controlled by a separate but related JSON structure and are documented in the missions docs.

Two topics are special:

  • TALK_DONE ends the dialogue immediately.
  • TALK_NONE goes to the previously talked about topic.

If npc has the following fields, the game will display a dialogue with the indicated topic instead of default topic.

Field Default topic ID Uses for…
talk_friend TALK_FRIEND Talk to a follower NPC
talk_radio TALK_RADIO Talk to a follower NPC with two way radios
talk_leader TALK_LEADER Talk to an NPC that have 5=NPCATT_LEAD
talk_stole_item TALK_STOLE_ITEM Talk to an NPC that have 18=NPCATT_RECOVER_GOODS
talk_wake_up TALK_WAKE_UP Talk to a sleeping NPC
talk_friend_guard TALK_FRIEND_GUARD Faction camp guard
talk_mug TALK_MUG see “success and failure” section
talk_stranger_aggressive TALK_STRANGER_AGGRESSIVE see “success and failure” section
talk_stranger_scared TALK_STRANGER_SCARED see “success and failure” section
talk_stranger_wary TALK_STRANGER_WARY see “success and failure” section
talk_stranger_friendly TALK_STRANGER_FRIENDLY see “success and failure” section
talk_stranger_neutral TALK_STRANGER_NEUTRAL see “success and failure” section

Validating Dialogues

Keeping track of talk topics and making sure that all the topics referenced in responses are defined, and all defined topics are referenced in a response or an NPC’s chat, is very tricky. There is a python script in tools/dialogue_validator.py that will map all topics to responses and vice versa. Invoke it with

python3 tools/dialogue_validator.py data/json/npcs/* data/json/npcs/Backgrounds/* data/json/npcs/refugee_center/*

If you are writing a mod with dialogue, you can add the paths to the mod’s dialogue files.

Customizing NPC speech

NPCs have dialogue depending on the situation. This dialogue can be customized in "type": "npc" json entries.

If npc has one of the following fields, the NPCs will speak the indicated message or snippet instead of the default message.

All messages can be used with snippets. Any %s included is automatically replaced by the game, with words depending on the message.

Case use example:

{
  "type":"npc",
  "...": "rest of fields go here",
  "<acknowledged>": "I gotcha fam",
  "<camp_food_thanks>": "<food_thanks_custom>"
},
{
  "type":"snippet",
  "category":"<food_thanks_custom>",
  "text": [
    "thanks for the grub",
    "thanks for the food!",
    "itadakimasu!"
  ]
}

For further information on snippets, see New Contributor Guide: Dialogue

Custom Entries

Field Default messages/snippets Used for…
<acknowledged> <acknowledged> see data/json/npcs/talk_tags.json
<camp_food_thanks> <camp_food_thanks> see data/json/npcs/talk_tags.json
<camp_larder_empty> <camp_larder_empty> see data/json/npcs/talk_tags.json
<camp_water_thanks> <camp_water_thanks> see data/json/npcs/talk_tags.json
<cant_flee> <cant_flee> see data/json/npcs/talk_tags.json
<close_distance> <close_distance> see data/json/npcs/talk_tags.json
<combat_noise_warning> <combat_noise_warning> see data/json/npcs/talk_tags.json
<danger_close_distance> <danger_close_distance> see data/json/npcs/talk_tags.json
<done_mugging> <done_mugging> see data/json/npcs/talk_tags.json
<far_distance> <far_distance> see data/json/npcs/talk_tags.json
<fire_bad> <fire_bad> see data/json/npcs/talk_tags.json
<fire_in_the_hole_h> <fire_in_the_hole_h> see data/json/npcs/talk_tags.json
<fire_in_the_hole> <fire_in_the_hole> see data/json/npcs/talk_tags.json
<general_danger_h> <general_danger_h> see data/json/npcs/talk_tags.json
<general_danger> <general_danger> see data/json/npcs/talk_tags.json
<heal_self> <heal_self> see data/json/npcs/talk_tags.json
<hungry> <hungry> see data/json/npcs/talk_tags.json
<im_leaving_you> <im_leaving_you> see data/json/npcs/talk_tags.json
<its_safe_h> <its_safe_h> see data/json/npcs/talk_tags.json
<its_safe> <its_safe> see data/json/npcs/talk_tags.json
<keep_up> <keep_up> see data/json/npcs/talk_tags.json
<kill_npc_h> <kill_npc_h> see data/json/npcs/talk_tags.json
<kill_npc> <kill_npc> see data/json/npcs/talk_tags.json
<kill_player_h> <kill_player_h> see data/json/npcs/talk_tags.json
<let_me_pass> <let_me_pass> see data/json/npcs/talk_tags.json
<lets_talk> <lets_talk> see data/json/npcs/talk_tags.json
<medium_distance> <medium_distance> see data/json/npcs/talk_tags.json
<monster_warning_h> <monster_warning_h> see data/json/npcs/talk_tags.json
<monster_warning> <monster_warning> see data/json/npcs/talk_tags.json
<movement_noise_warning> <movement_noise_warning> see data/json/npcs/talk_tags.json
<need_batteries> <need_batteries> see data/json/npcs/talk_tags.json
<need_booze> <need_booze> see data/json/npcs/talk_tags.json
<need_fuel> <need_fuel> see data/json/npcs/talk_tags.json
<no_to_thorazine> <no_to_thorazine> see data/json/npcs/talk_tags.json
<run_away> <run_away> see data/json/npcs/talk_tags.json
<speech_warning> <speech_warning> see data/json/npcs/talk_tags.json
<thirsty> <thirsty> see data/json/npcs/talk_tags.json
<wait> <wait> see data/json/npcs/talk_tags.json
<warn_sleep> <warn_sleep> see data/json/npcs/talk_tags.json
<yawn> <yawn> see data/json/npcs/talk_tags.json
<yes_to_lsd> <yes_to_lsd> see data/json/npcs/talk_tags.json
snip_bleeding My %s is bleeding! The NPC is bleeding from their %s.
snip_bleeding_badly My %s is bleeding badly! The NPC is bleeding badly from their %s.
snip_wound_bite The bite wound on my %s looks bad. The NPC’s %s was badly bitten.
snip_wound_infected My %s wound is infected… The NPC’s wound on their %s is infected.
snip_lost_blood I've lost lot of blood. The NPC has lost lot of blood.
snip_bye Bye. Say at the end of a conversation.
snip_consume_eat Thanks, that hit the spot. You gave them some food and the NPC ate it.
snip_consume_cant_accept I don't <swear> trust you enough to eat THIS… You gave them some food but the NPC didn’t accept it.
snip_consume_cant_consume It doesn't look like a good idea to consume this… You gave them some food but the NPC wouldn’t eat it.
snip_consume_rotten This is rotten! I won't eat that. You gave them some food but the NPC wouldn’t eat it because it’s rotten.
snip_consume_med Thanks, I feel better already. You gave them a drug and the NPC swallowed it.
snip_consume_use_med Thanks, I used it. You gave them a drug and the NPC used it.
snip_consume_need_item I need a %s to consume that! You gave them a drug but the NPC couldn’t use it without a %s (e.g., a syringe).
snip_consume_nocharge It doesn't look like a good idea to consume this… You gave them a drug but the NPC couldn’t use it because of a lack of charges.
snip_give_cancel Changed your mind? You started to give them an item but hit the escape key.
snip_give_carry Thanks, I'll carry that now. You gave them an item and the NPC took it.
snip_give_nope Nope. You gave them an item but the NPC couldn’t take it. First says.
snip_give_carry_cant I have no space to store it. You gave an item but NPC couldn’t take because no space.
snip_give_carry_cant_few_space I can only store %s %s more. You gave an item but NPC couldn’t take because few spaces.
snip_give_carry_cant_no_space …or to store anything else for that matter. You gave an item but NPC couldn’t take because no spaces.
snip_give_carry_too_heavy It is too heavy for me to carry. You gave an item but NPC couldn’t take because too heavy.
snip_give_dangerous Are you <swear> insane!? You gave a dangerous item like active grenade.
snip_give_to_hallucination No thanks, I'm good. You gave an item but NPC couldn’t take because he/she was in hallucination.
snip_give_wield Thanks, I'll wield that now. You gave a weapon and NPC wield it.
snip_give_weapon_weak My current weapon is better than this.\n You gave a weapon but NPC didn’t take it because it was weaker than current weapons.
snip_heal_player Hold still %s, I'm coming to help you. NPC try to heal you.
snip_pulp_zombie Hold on, I want to pulp that %s. NPC try to pulp %s.
snip_radiation_sickness I'm suffering from radiation sickness… NPC are in radiation sickness.
snip_wear Thanks, I'll wear that now. You gave a clothes or armor and NPC wear it.

Special Custom Entries

Certain entries like the snippets above are taken from the game state as opposed to JSON; they are found in the npctalk function parse_tags. They are as follows:

Field Used for…
<yrwp> displays avatar’s wielded item
<mywp> displays npc’s wielded item
<u_name> displays avatar’s name
<npc_name> displays npc’s name
<ammo> displays avatar’s ammo
<current_activity> displays npc’s current activity
<punc> displays a random punctuation from: ., …, !
<mypronoun> displays npc’s pronoun
<topic_item> referenced item
<topic_item_price> referenced item unit price
<topic_item_my_total_price> TODO Add
<topic_item_your_total_price> TODO Add
<interval> displays the time remaining until restock
<u_val:VAR> The user variable VAR
<npc_val:VAR> The npc variable VAR
<context_val:VAR> The context variable VAR
<global_val:VAR> The global variable VAR
<item_name:ID> The name of the item from ID
<item_description:ID> The description of the item from ID
<trait_name:ID> The name of the trait from ID
<trait_description:ID> The description of the trait from ID
<spell_name:ID> The description of the name from ID
<spell_description:ID> The description of the trait from ID

item_name and similar tags, that parse the text out of the id, are able to parse the tags of variables, so it is possible to use <item_name:<global_val:VAR>>

Talk Topics

Each topic consists of:

  1. a topic id (e.g. TALK_ARSONIST)
  2. a dynamic line, spoken by the NPC.
  3. an optional list of effects that occur when the NPC speaks the dynamic line
  4. a list of responses that can be spoken by the player character.
  5. a list of repeated responses that can be spoken by the player character, automatically generated if the player or NPC has items in a list of items.

One can specify new topics in json. It is currently not possible to define the starting topic, so you have to add a response to some of the default topics (e.g. TALK_STRANGER_FRIENDLY or TALK_STRANGER_NEUTRAL) or to topics that can be reached somehow.

Format:

{
  "type": "talk_topic",
  "id": "TALK_ARSONIST",
  "dynamic_line": "What now?",
  "responses": [
    {
      "text": "I don't know either",
      "topic": "TALK_DONE"
    }
  ],
  "replace_built_in_responses": true
}

type

Must always be there and must always be "talk_topic".

"insert_before_standard_exits"

For mod usage to insert dialogue above the TALK_DONE and TALK_NONE lines. Defaults to false.

  {
    "id": "TALK_REFUGEE_Draco_1a",
    "type": "talk_topic",
    "insert_before_standard_exits": true,
    "responses": [ { "text": "Have you seen anything that could help me?", "topic": "TALK_REFUGEE_Draco_changeling_breadcrumb" } ]
  }

id

The topic id can be one of the built-in topics or a new id. However, if several talk topics in json have the same id, the last topic definition will override the previous ones.

The topic id can also be an array of strings. This is loaded as if several topics with the exact same content have been given in json, each associated with an id from the id, array. Note that loading from json will append responses and, if defined in json, override the dynamic_line and the replace_built_in_responses setting. This allows adding responses to several topics at once.

This example adds the “I’m going now!” response to all the listed topics.

{
    "type": "talk_topic",
    "id": [ "TALK_ARSONIST", "TALK_STRANGER_FRIENDLY", "TALK_STRANGER_NEUTRAL" ],
    "dynamic_line": "What now?",
    "responses": [
        {
            "text": "I'm going now.",
            "topic": "TALK_DONE"
        }
    ]
}

dynamic_line

The dynamic_line is the line spoken by the NPC. It is optional. If it is not defined and the topic has the same id as a built-in topic, the dynamic_line from that built-in topic will be used. Otherwise the NPC will say nothing. See the chapter about Dynamic Lines below for more details.

speaker_effect

The speaker_effect is an object or array of effects that will occur after the NPC speaks the dynamic_line, no matter which response the player chooses. See the chapter about Speaker Effects below) for more details.

response

The responses entry is an array with possible responses. It must not be empty. Each entry must be a response object. See the chapter about Responses below for more details.

replace_built_in_responses

replace_built_in_responses is an optional boolean that defines whether to dismiss the built-in responses for that topic (default is false). If there are no built-in responses, this won’t do anything. If true, the built-in responses are ignored and only those from this definition in the current json are used. If false, the responses from the current json are used along with the built-in responses (if any).

Dynamic Lines

A dynamic line can either be a simple string, or an complex object, or an array with dynamic_line entries. If it’s an array, an entry will be chosen randomly every time the NPC needs it. Each entry has the same probability.

Example:

"dynamic_line": [
  "generic text",
  {
    "npc_female": [ "text1", "text2", "text3" ],
    "npc_male": { "u_female": "text a", "u_male": "text b" }
  }
]

A complex dynamic_line usually contains several dynamic_line entry and some condition that determines which is used. If dynamic lines are not nested, they are processed in the order of the entries below. The possible types of lines follow.

In all cases, npc_ refers to the NPC, and u_ refers to the player. Optional lines do not have to be defined, but the NPC should always have something to say. Entries are always parsed as dynamic_line and can be nested.

Several lines joined together

The dynamic line is a list of dynamic lines, all of which are displayed. The dynamic lines in the list are processed normally.

{
  "concatenate": [
    {
      "npc_male": true,
      "yes": "I'm a man.",
      "no": "I'm a woman."
    },
    "  ",
    {
      "u_female": true,
      "no": "You're a man.",
      "yes": "You're a woman."
    }
  ]
}

A line to be translated with gender context

The line is to be given a gender context for the NPC, player, or both, to aid translation in languages where that matters. For example:

{
  "gendered_line": "Thank you.",
  "relevant_genders": [ "npc" ]
}

(“Thank you” is different for male and female speakers in e.g. Portuguese).

Valid choices for entries in the "relevant_genders" list are "npc" and "u".

A randomly selected hint

The dynamic line will be randomly chosen from the hints snippets.

{
  "give_hint": true
}

A list of potential faction camp sites

The dynamic line will list all of the possible starting sites for faction camps.

{
  "list_faction_camp_sites": true
}

Based on a previously generated reason

The dynamic line will be chosen from a reason generated by an earlier effect. The reason will be cleared. Use of it should be gated on the "has_reason" condition.

{
  "has_reason": { "use_reason": true },
  "no": "What is it, boss?"
}

Based on any dialogue condition

The dynamic line will be chosen based on whether a single dialogue condition is true or false. Dialogue conditions cannot be chained via "and", "or", or "not". If the condition is true, the "yes" response will be chosen and otherwise the "no" response will be chosen. Both the '"yes" and "no" responses are optional. Simple string conditions may be followed by "true" to make them fields in the dynamic line dictionary, or they can be followed by the response that will be chosen if the condition is true and the "yes" response can be omitted.

{
  "npc_need": "fatigue",
  "level": "TIRED",
  "no": "Just few minutes more...",
  "yes": "Make it quick, I want to go back to sleep."
}
{
  "npc_aim_rule": "AIM_PRECISE",
  "no": "*will not bother to aim at all.",
  "yes": "*will take time and aim carefully."
}
{
  "u_has_item": "india_pale_ale",
  "yes": "<noticedbooze>",
  "no": "<neutralchitchat>"
}
{
  "days_since_cataclysm": 30,
  "yes": "Now, we've got a moment, I was just thinking it's been a month or so since... since all this, how are you coping with it all?",
  "no": "<neutralchitchat>"
}
{
  "is_day": "Sure is bright out.",
  "no": {
    "u_male": true,
      "yes": "Want a beer?",
      "no": "Want a cocktail?"
  }
}

Speaker Effects

The speaker_effect entry contains dialogue effects that occur after the NPC speaks the dynamic_line but before the player responds and regardless of the player response. Each effect can have an optional condition, and will only be applied if the condition is true. Each speaker_effect can also have an optional sentinel, which guarantees the effect will only run once.

Format:

"speaker_effect": {
  "sentinel": "...",
  "condition": "...",
  "effect": "..."
}

or:

"speaker_effect": [
  {
    "sentinel": "...",
    "condition": "...",
    "effect": "..."
  },
  {
    "sentinel": "...",
    "condition": "...",
    "effect": "..."
  }
]

The sentinel can be any string, but sentinels are unique to each TALK_TOPIC. If there are multiple speaker_effects within the TALK_TOPIC, they should have different sentinels. Sentinels are not required, but since the speaker_effect will run every time the dialogue returns to the TALK_TOPIC, they are highly encouraged to avoid inadvertently repeating the same effects.

The effect can be any legal effect, as described below. The effect can be a simple string, object, or an array of strings and objects, as normal for objects.

The optional condition can be any legal condition, as described below. If a condition is present, the effect will only occur if the condition is true.

Speaker effects are useful for setting status variables to indicate that player has talked to the NPC without complicating the responses with multiple effect variables. They can also be used, with a sentinel, to run a mapgen_update effect the first time the player hears some dialogue from the NPC.

Responses

A response contains at least a text, which is display to the user and “spoken” by the player character (its content has no meaning for the game) and a topic to which the dialogue will switch to. It can also have a trial object which can be used to either lie, persuade or intimidate the NPC, see below for details. There can be different results, used either when the trial succeeds and when it fails.

Format:

{
  "text": "I, the player, say to you...",
  "condition": "...something...",
  "trial": {
    "type": "PERSUADE",
    "difficulty": 10
  },
  "success": {
    "topic": "TALK_DONE",
    "effect": "...",
    "opinion": {
      "trust": 0,
      "fear": 0,
      "value": 0,
      "anger": 0,
      "owed": 0,
      "favors": 0
    }
  },
  "failure": {
    "topic": "TALK_DONE"
  }
}

Alternatively a short format:

{
  "text": "I, the player, say to you...",
  "effect": "...",
  "topic": "TALK_WHATEVER"
}

The short format is equivalent to (an unconditional switching of the topic, effect is optional):

{
  "text": "I, the player, say to you...",
  "trial": {
    "type": "NONE"
  },
  "success": {
    "effect": "...",
    "topic": "TALK_WHATEVER"
  }
}

When using a conditional you can specify the response to still appear but be marked as unavailable. This can be done by adding a failure_explanation or failure_topic in the bellow example if the condition fails *Didn't have enough: I, the player, say to you... will be what appears in the responses, and if selected it will instead go to TALK_EXPLAIN_FAILURE and wont trigger the other effects:

{
  "condition": "...something...",
  "failure_explanation": "Didn't have enough",
  "failure_topic": "TALK_EXPLAIN_FAILURE",
  "text": "I, the player, say to you...",
  "effect": "...",
  "topic": "TALK_WHATEVER"
}

text

Will be shown to the user, no further meaning.

Text boxes; dialogue in general is a convenient space to sprinkle in descriptive text, something that isn’t necessarily being said by any interlocutor but something the player character, npc or speaking entity express, do or generally interact with given a context there are many ways to present this, ultimately is up to the writer, and their preferred style.

Currently you may add a & as the first character in dialogue, this deletes quotation round the output text, denotes the descriptive nature of the displayed text, use \" escaped double quotes to indicate the start of actual dialogue.

truefalsetext

May be used in place of text. The player will have one response text if a condition is true, and another if it is false, but the same trial for either line. condition, true, and false are all mandatory.

{
  "truefalsetext": {
    "condition": { "u_has_cash": 800 },
    "true": "I may have the money, I'm not giving you any.",
    "false": "I don't have that money."
  },
  "topic": "TALK_WONT_PAY"
}

topic

topic defines which topic the dialogue will switch to, usually specified by giving its id.

topic can also be a single topic object (the type member is not required here):

"success": {
  "topic": {
    "id": "TALK_NEXT",
    "dynamic_line": "...",
    "responses": [
    ]
  }
}

effect

effect is a function that is executed after choosing the response, see Dialogue Effects below for details.

Trials

A trial object can be used to attempt to lie to, persuade or intimidate the NPC. Different outcomes can be defined for use depending on whether the trial succeeds or fails.

trial

Optional, if not defined, "NONE" is used. Otherwise one of "NONE", "LIE", "PERSUADE", "INTIMIDATE", or "CONDITION". If "NONE" is used, the failure object is not read, otherwise it’s mandatory.

The difficulty is only required if type is not "NONE" or "CONDITION" and, for most trials, specifies the success chance in percent (it is however modified by various things like mutations). Higher difficulties are easier to pass. "SKILL_CHECK" trials are unique, and use the difficulty as a flat comparison.

An optional mod array takes any of the following modifiers and increases the difficulty by the NPC’s opinion of your character or personality trait for that modifier multiplied by the value: "ANGER", "FEAR", "TRUST", "VALUE", "AGGRESSION", "ALTRUISM", "BRAVERY", "COLLECTOR". The special "POS_FEAR" modifier treats NPC’s fear of your character below 0 as though it were 0. The special "TOTAL" modifier sums all previous modifiers and then multiplies the result by its value and is used when setting the owed value.

"CONDITION" trials take a mandatory condition instead of difficulty. The success object is chosen if the condition is true and the failure is chosen otherwise.

"SKILL_CHECK" trials check the user’s level in a skill, whose ID is read from the string object skill_required. The success object is chosen if the skill level is equal to or greater than difficulty, and failure is chosen otherwise.

Sample trials:

"trial": { "type": "PERSUADE", "difficulty": 0, "mod": [ [ "TRUST", 3 ], [ "VALUE", 3 ], [ "ANGER", -3 ] ] }
"trial": { "type": "INTIMIDATE", "difficulty": 20, "mod": [ [ "FEAR", 8 ], [ "VALUE", 2 ], [ "TRUST", 2 ], [ "BRAVERY", -2 ] ] }
"trial": { "type": "CONDITION", "condition": { "npc_has_trait": "FARMER" } }
"trial": { "type": "SKILL_CHECK", "difficulty": 3, "skill_required": "swimming" }

success and failure

The success and failure objects define the outcome, depending on the result of the trial. Both objects have the same structure; the failure object is used if the trial fails, the success object is used otherwise.

Opinion Changes

opinion

opinion is optional, if given it defines how the NPC’s opinion of your character will change.

trust, value, fear, and anger are optional fields inside the opinion object, each specifying a numeric value (defaults to 0). The given values are added to the opinion of the NPC.

The opinion of the NPC affects several aspects of the interaction with NPCs:

  • Higher trust makes it easier to lie and persuade, and it usually a good thing.
  • Higher fear makes it easier to intimidate, but the NPC may flee from you (and will not talk to you).
  • Higher value makes it easier to persuade them and to give them orders, it’s a kind of a friendship indicator.
  • High anger value (about 20 points more than fear, but this also depends on the NPCs personality) makes the NPC hostile and is usually a bad thing. The combination of fear and trust decide together with the personality of the NPC the initial talk topic ("TALK_MUG", "TALK_STRANGER_AGGRESSIVE", "TALK_STRANGER_SCARED", "TALK_STRANGER_WARY", "TALK_STRANGER_FRIENDLY", or "TALK_STRANGER_NEUTRAL").

For the actual usage of that data, search the source code for "op_of_u".

Example opinions

{ "effect": "follow", "opinion": { "trust": 1, "value": 1 }, "topic": "TALK_DONE" }
{ "topic": "TALK_DENY_FOLLOW", "effect": "deny_follow", "opinion": { "fear": -1, "value": -1, "anger": 1 } }

mission_opinion

Similar to opinion, but adjusts the NPC’s opinion of your character according to the mission value. The NPC’s opinion is modified by the value of the current mission divided by the value of the keyword.

Response Availability

condition

This is an optional condition which can be used to prevent the response under certain circumstances. If not defined, it defaults to always true. If the condition is not met, the response is not included in the list of possible responses. For possible content, see Dialogue Conditions below for details.

switch and default

The optional boolean keys “switch” and “default” are false by default. Only the first response with "switch": true, "default": false, and a valid condition will be displayed, and no other responses with "switch": true will be displayed. If no responses with "switch": true and "default": false are displayed, then any and all responses with "switch": true and "default": true will be displayed. In either case, all responses that have "switch": false (whether or not they have "default": true is set) will be displayed as long their conditions are satisfied.

Example:

"responses": [
  { "text": "You know what, never mind.", "topic": "TALK_NONE" },
  { "text": "How does 5 Ben Franklins sound?",
    "topic": "TALK_BIG_BRIBE", "condition": { "u_has_cash": 500 }, "switch": true },
  { "text": "I could give you a big Grant.",
    "topic": "TALK_BRIBE", "condition": { "u_has_cash": 50 }, "switch": true },
  { "text": "Lincoln liberated the slaves, what can he do for me?",
    "topic": "TALK_TINY_BRIBE", "condition": { "u_has_cash": 5 }, "switch": true, "default": true },
  { "text": "Maybe we can work something else out?", "topic": "TALK_BRIBE_OTHER",
    "switch": true, "default": true },
  { "text": "Gotta go!", "topic": "TALK_DONE" }
]

The player will always have the option to return to a previous topic or end the conversation, and will otherwise have the option to give a $500, $50, or $5 bribe if they have the funds. If they don’t have at least $50, they will also have the option to provide some other bribe.

Repeat Responses

Repeat responses are responses that should be added to the response list multiple times, once for each instance of an item.

A repeat response has the following format:

{
  "for_item": [
    "jerky", "meat_smoked", "fish_smoked", "cooking_oil", "cooking_oil2", "cornmeal", "flour",
    "fruit_wine", "beer", "sugar"
  ],
  "response": { "text": "Delivering <topic_item>.", "topic": "TALK_DELIVER_ASK" }
}

"response" is mandatory and must be a standard dialogue response, as described above. "switch" is allowed in repeat responses and works normally.

One of "for_item" or "for_category", and each can either be a single string or list of items or item categories. The response is generated for each item in the list in the player or NPC’s inventory.

"is_npc" is an optional bool value, and if it is present, the NPC’s inventory list is checked. By default, the player’s inventory list is checked.

"include_containers" is an optional bool value, and if it is present, items containing an item will generate separate responses from the item itself.

Dialogue State

Variables and information relevant to the current dialogue can be tracked using context variables. Accessing these is discussed further in variable object. The main thing that makes context variables special however is that they are only relevant to the current dialogue and any child dialogue / effects. When the dialogue or effect ends any context variables defined inside go out of scope (stop existing).

Dialogue Effects

The effect field of speaker_effect or a response can be any of the following effects. Multiple effects should be arranged in a list and are processed in the order listed.

Missions

Effect Description
assign_mission Assigns a previously selected mission to your character.
mission_success Resolves the current mission successfully.
mission_failure Resolves the current mission as a failure.
finish_mission Resolves a specific mission defined by ID
clear_mission Clears the mission from the your character’s assigned missions.
mission_reward Gives the player the mission’s reward.

Stats / Morale

Effect Description
lesser_give_aid Removes bleeding from your character’s body and heals 5-15 HP of injury on each of your character’s body parts.
lesser_give_aid_all Performs lesser_give_aid on each of your character’s NPC allies in range.
give_aid Removes all bites, infection, and bleeding from your character’s body and heals 10-25 HP of injury on each of your character’s body parts.
give_aid_all Performs give_aid on each of your character’s NPC allies in range.
buy_haircut Gives your character a haircut morale boost for 12 hours.
buy_shave Gives your character a shave morale boost for 6 hours.
morale_chat Gives your character a pleasant conversation morale boost for 6 hours.
player_weapon_away Makes your character put away (unwield) their weapon.
player_weapon_drop Makes your character drop their weapon.

Character effects / Mutations

See EFFECT_ON_CONDITION.md, #Character effects

Trade / Items

Effect Description
start_trade Opens the trade screen and allows trading with the NPC.
give_equipment Allows your character to select items from the NPC’s inventory and transfer them to your inventory.
npc_gets_item Allows your character to select an item from your character’s inventory and transfer it to the NPC’s inventory. The NPC will not accept it if they do not have space or weight to carry it, and will set a reason that can be referenced in a future dynamic line with "use_reason".
npc_gets_item_to_use Allow your character to select an item from your character’s inventory and transfer it to the NPC’s inventory. The NPC will attempt to wield it and will not accept it if it is too heavy or is an inferior weapon to what they are currently using, and will set a reason that can be referenced in a future dynamic line with "use_reason".
u_spawn_item: string or variable object, (optional count: int or variable object), (optional container: string or variable object), (optional use_item_group: bool), (optional suppress_message: bool), (optional flags: array of string or variable object) Your character gains the item or count copies of the item, contained in container if specified. If used in an NPC conversation the items are said to be given by the NPC. If a variable item is passed for the name an item of the type contained in it will be used. If use_item_group is true (defaults to false) it will instead pull an item from the item group given. If suppress_message is true (defaults to false) no message will be shown. If force_equip is true (defaults to false) characters will equip the items if they can. The item will have all the flags from the array flags.
u_buy_item: string or variable object, cost: int or variable object, (optional count: int or variable object), (optional container: string or variable object), (optional true_eocs: eocs_array), (optional false_eocs: eocs_array), (optional use_item_group: bool), (optional suppress_message: bool), (optional flags: array of string or variable object) The NPC will sell your character the item or count copies of the item, contained in container, and will subtract cost from op_of_u.owed. If the op_o_u.owed is less than cost, the trade window will open and the player will have to trade to make up the difference; the NPC will not give the player the item unless cost is satisfied. If use_item_group is true (defaults to false) it will instead pull an item from the item group given. If suppress_message is true (defaults to false) no message will be shown. The item will have all the flags from the array flags.
u_sell_item: string or variable object, (optional cost: int or variable object), (optional count: string or variable object), (optional true_eocs: eocs_array), (optional false_eocs: eocs_array) Your character will give the NPC the item or count copies of the item, and will add cost to the NPC’s op_of_u.owed if specified.
If cost isn’t present, the your character gives the NPC the item at no charge.
This effect will fail if you do not have at least count copies of the item, so it should be checked with. If the item is sold, then all of the effect_on_conditions in true_eocs are run, otherwise all the effect_on_conditions in false_eocs are run.
u_bulk_trade_accept, npc_bulk_trade_accept, u_bulk_trade_accept, npc_bulk_trade_accept: int or variable object Only valid after a repeat_response. The player trades all instances of the item from the repeat_response with the NPC. For u_bulk_trade_accept, the player loses the items from their inventory and gains the same value of the NPC’s faction currency; for npc_bulk_trade_accept, the player gains the items from the NPC’s inventory and loses the same value of the NPC’s faction currency. If there is remaining value, or the NPC doesn’t have a faction currency, the remainder goes into the NPC’s op_of_u.owed. If quantity is specified only that many items/charges will be moved.
u_bulk_donate, npc_bulk_donate or u_bulk_donate, npc_bulk_donate: int or variable object Only valid after a repeat_response. The player or NPC transfers all instances of the item from the repeat_response. For u_bulk_donate, the player loses the items from their inventory and the NPC gains them; for npc_bulk_donate, the player gains the items from the NPC’s inventory and the NPC loses them. If a value is specified only that many items/charges will be moved.
u_spend_cash: int or variable object, (optional true_eocs: eocs_array), (optional false_eocs: eocs_array) Remove an amount from your character’s cash. Negative values means your character gains cash. deprecated NPCs should not deal in e-cash anymore, only personal debts and items. If the cash is spent, then all of the effect_on_conditions in true_eocs are run, otherwise all the effect_on_conditions in false_eocs are run.
add_debt: mod_list Increases the NPC’s debt to the player by the values in the mod_list.
The following would increase the NPC’s debt to the player by 1500x the NPC’s altruism and 1000x the NPC’s opinion of the player’s value: { "effect": { "add_debt": [ [ "ALTRUISM", 3 ], [ "VALUE", 2 ], [ "TOTAL", 500 ] ] } }
u_consume_item, npc_consume_item: string or variable object, (optional count: int or variable object), (optional charges: int or variable object), (optional popup: popup_bool) You or the NPC will delete the item or count copies of the item or charges charges of the item from their inventory.
This effect will fail if the you or NPC does not have at least count copies of the item or charges charges of the item, so it should be checked with u_has_items or npc_has_items.
If popup_bool is true, u_consume_item will show a message displaying the character giving the items to the NPC. It defaults to false if not defined, and has no effect when used in npc_consume_item.
u_remove_item_with, npc_remove_item_with: string or variable object You or the NPC will delete any instances of item in inventory.
This is an unconditional remove and will not fail if you or the NPC does not have the item.
u_buy_monster: string or variable object, (optional cost: int or variable object), (optional count: int or variable object), (optional name: string or variable object), (optional pacified: pacified_bool), (optional true_eocs: eocs_array), (optional false_eocs: eocs_array) The NPC will give your character count (default 1) instances of the monster as pets and will subtract cost from op_of_u.owed if specified. If the op_o_u.owed is less than cost, the trade window will open and the player will have to trade to make up the difference; the NPC will not give the player the item unless cost_num is satisfied.
If cost isn’t present, the NPC gives your character the item at no charge.
If name is specified the monster(s) will have the specified name. If pacified_bool is set to true, the monster will have the pacified effect applied. If the monster is sold, then all of the effect_on_conditions in true_eocs are run, otherwise all the effect_on_conditions in false_eocs are run.

Behavior / AI

Effect Description
assign_guard Makes the NPC into a guard. If allied and at a camp, they will be assigned to that camp.
stop_guard Releases the NPC from their guard duty (also see assign_guard). Friendly NPCs will return to following.
start_camp The NPC will start a faction camp with the player.
wake_up Wakes up sleeping, but not sedated, NPCs.
reveal_stats Reveals the NPC’s stats, based on the player’s skill at assessing them.
end_conversation Ends the conversation and makes the NPC ignore you from now on.
insult_combat Ends the conversation and makes the NPC hostile, adds a message that character starts a fight with the NPC.
hostile Makes the NPC hostile and ends the conversation.
flee Makes the NPC flee from your character.
follow Makes the NPC follow your character, joining the “Your Followers” faction.
leave Makes the NPC leave the “Your Followers” faction and stop following your character.
follow_only Makes the NPC follow your character without changing factions.
stop_following Makes the NPC stop following your character without changing factions.
npc_thankful Makes the NPC positively inclined toward your character.
drop_weapon Makes the NPC drop their weapon.
stranger_neutral Changes the NPC’s attitude to neutral.
start_mugging The NPC will approach your character and steal from your character, attacking if your character resists.
lead_to_safety The NPC will gain the LEAD attitude and give your character the mission of reaching safety.
start_training The NPC will train your character in a skill or martial art. NOTE: the code currently requires that you initiate training by directing the player through "topic": "TALK_TRAIN" where the thing to be trained is selected. Initiating training outside of “TALK_TRAIN” will give an error.
start_training_npc The NPC will accept training from the player in a skill or martial art.
start_training_seminar Opens a dialog to select which characters will participate in the training seminar hosted by this NPC.
companion_mission: role_string The NPC will offer you a list of missions for your allied NPCs, depending on the NPC’s role.
basecamp_mission The NPC will offer you a list of missions for your allied NPCs, depending on the local basecamp.
bionic_install The NPC installs a bionic from your character’s inventory onto your character, using very high skill, and charging you according to the operation’s difficulty.
bionic_remove The NPC removes a bionic from your character, using very high skill, and charging you according to the operation’s difficulty.
npc_class_change: string or variable object Change the NPC’s class to the new value.
npc_faction_change: string or variable object Change the NPC’s faction membership to the new value.
u_faction_rep: int or variable object Increases your reputation with the NPC’s current faction, or decreases it if the value is negative.
toggle_npc_rule: string or variable object Toggles the value of a boolean NPC follower AI rule such as "use_silent" or "allow_bash"
set_npc_rule: string or variable object Sets the value of a boolean NPC follower AI rule such as "use_silent" or "allow_bash"
clear_npc_rule: string or variable object Clears the value of a boolean NPC follower AI rule such as "use_silent" or "allow_bash"
set_npc_engagement_rule: string or variable object Sets the NPC follower AI rule for engagement distance to the value of rule_string.
set_npc_aim_rule: string or variable object Sets the NPC follower AI rule for aiming speed to the value of rule_string.
npc_die The NPC will die at the end of the conversation.
u_set_goal, npc_set_goal:assign_mission_target_object, (optional true_eocs: eocs_array), (optional false_eocs: eocs_array) The NPC will walk to assign_mission_target_object. See the missions docs for assign_mission_target parameters. If the goal is assigned, then all of the effect_on_conditions in true_eocs are run, otherwise all the effect_on_conditions in false_eocs are run.
u_set_guard_pos,npc_set_guard_pos : variable object, (optional unique_id: bool) Set the NPC’s guard pos to the contents of _set_guard_pos. If the NPC has the RETURN_TO_START_POS trait then when they are idle they will attempt to move to this position. If unique_id(defaults to false) is true then the NPC’s unique_id will be added as a prefix to the variables name. For example a guard with unique_id = GUARD1 would check the variable GUARD1_First in the following json statement: { "u_set_guard_pos": { "global_val": "_First" }, "unique_id": true }

Map Updates

Map Updates

General

Map Updates

Deprecated

Effect Description
deny_follow
deny_lead
deny_train
deny_personal_info		 Sets the appropriate effect on the NPC for a few hours.
These are deprecated in favor of the more flexible npc_add_effect described above.

Sample effects 

{ "topic": "TALK_EVAC_GUARD3_HOSTILE", "effect": [ { "u_faction_rep": -15 }, { "npc_change_faction": "hells_raiders" } ] }
{ "text": "Let's trade then.", "effect": "start_trade", "topic": "TALK_EVAC_MERCHANT" }
{ "text": "Show me what needs to be done at the camp.", "topic": "TALK_DONE", "effect": "basecamp_mission", "condition": { "npc_at_om_location": "FACTION_CAMP_ANY" } }
{ "text": "Do you like it?", "topic": "TALK_EXAMPLE", "effect": [ { "u_add_effect": "concerned", "duration": 600 }, { "npc_add_effect": "touched", "duration": 3600 }, { "u_add_effect": "empathetic", "duration": "PERMANENT" } ] }

Dialogue Conditions

Conditions

Player Only conditions

Condition Type Description
"u_has_mission" string or variable object true if the mission is assigned to the player character.
"u_has_cash" int or variable object true if the player character has at least u_has_cash cash available. Deprecated Previously used to check if the player could buy something, but NPCs shouldn’t use e-cash for trades anymore.
"u_are_owed" int or variable object true if the NPC’s op_of_u.owed is at least u_are_owed. Can be used to check if the player can buy something from the NPC without needing to barter anything.
"u_has_camp" simple string true is the player has one or more active base camps.
"u_has_faction_trust" int or variable object true if the player character has at least the given amount of trust with the speaker’s faction.

Player and NPC interaction conditions

Condition Type Description
"at_safe_space" or "u_at_safe_space" or "npc_at_safe_space" simple string true if the only monsters present in the talker’s OMT are in a monster group with is_safe. False otherwise.
"has_assigned_mission" simple string true if the player character has exactly one mission from the NPC. Can be used for texts like “About that job…”.
"has_many_assigned_missions" simple string true if the player character has several mission from the NPC (more than one). Can be used for texts like “About one of those jobs…” and to switch to the "TALK_MISSION_LIST_ASSIGNED" topic.
"has_no_available_mission" or "npc_has_no_available_mission" or "u_has_no_available_mission" simple string true if u or the NPC has no jobs available for the player character.
"has_available_mission" or "u_has_available_mission" or "npc_has_available_mission" simple string true if u or the NPC has one job available for the player character.
"has_many_available_missions" simple string true if the NPC has several jobs available for the player character.
"mission_goal" or "npc_mission_goal" or "u_mission_goal" string or variable object true if u or the NPC’s current mission has the same goal as mission_goal.
"mission_complete" or "npc_mission_complete" or "u_mission_complete" simple string true if u or the NPC has completed the other’s current mission.
"mission_incomplete" or "npc_mission_incomplete" or "u_mission_incomplete" simple string true if u or the NPC hasn’t completed the other’s current mission.
"mission_failed" or "npc_mission_failed" or "u_mission_failed" simple string true if u or the NPC has failed the other’s current mission.
"mission_has_generic_rewards" simple string true if the NPC’s current mission is flagged as having generic rewards.
"npc_service" int or variable object true if the NPC does not have the "currently_busy" effect and the player character has at least npc_service cash available. Useful to check if the player character can hire an NPC to perform a task that would take time to complete. Functionally, this is identical to "and": [ { "not": { "npc_has_effect": "currently_busy" } }, { "u_has_cash": service_cost } ]
"npc_allies" int or variable object true if the player character has at least npc_allies number of NPC allies within the reality bubble.
"npc_allies_global" int or variable object true if the player character has at least npc_allies_global number of NPC allies anywhere.
"is_by_radio" simple string true if the player is talking to the NPC over a radio.
"u_available" or "npc_available" simple string true if u or the NPC does not have effect "currently_busy".
"u_following" or "npc_following" simple string true if u or the NPC is following the player character.
"u_friend" or "npc_friend" simple string true if u or the NPC is friendly to the player character.
"u_hostile" or "npc_hostile" simple string true if u or the NPC is an enemy of the player character.
"u_train_skills" or "npc_train_skills" simple string true if u or the NPC has one or more skills with more levels than the player.
"u_train_styles" or "npc_train_styles" simple string true if u or the NPC knows one or more martial arts styles that the player does not know.
"u_has_class" or "npc_has_class" string or variable object true if u or the NPC’s class id matches the provided string (e.g. NC_BANDIT_TRADER).
"u_near_om_location" or "npc_near_om_location", (optional range : int or variable object ) string or variable object same as at_om_location except it checks in a square stretching from the character range OMT’s. NOTE: can only check OMT’s in the reality bubble (maximum of ~2 OMTs distance from player’s position)
"u_aim_rule" or "npc_aim_rule" string or variable object true if u or the NPC follower AI rule for aiming matches the string.
"u_engagement_rule" or "npc_engagement_rule" string or variable object true if u or the NPC follower AI rule for engagement matches the string.
"u_cbm_reserve_rule" or "npc_cbm_reserve_rule" string or variable object true if u or the NPC follower AI rule for cbm, reserve matches the string.
"u_cbm_recharge_rule" or "npc_cbm_recharge_rule" string or variable object true if u or the NPC follower AI rule for cbm recharge matches the string.
"u_rule" or "npc_rule" string or variable object true if u or the NPC follower AI rule for that matches string is set.
"u_override" or "npc_override" string or variable object true if u or the NPC has an override for the string.
"has_pickup_list" or "u_has_pickup_list" or "npc_has_pickup_list" simple string true if u or the NPC has a pickup list.
"roll_contested"", difficulty: int or variable object, (optional die_size : int or variable object ) int expression Compares a roll against a difficulty. Returns true if a random number between 1 and die_size (defaults to 10) plus the integer expression is greater than difficulty. For example { “u_roll_contested”: { “u_val”: “strength” }, “difficulty”: 6 } will return whether a random number between 1 and 10 plus strength is greater than 6.

NPC only conditions

Condition Type Description
"npc_role_nearby" string or variable object true if there is an NPC with the same companion mission role as npc_role_nearby within 100 tiles.
"has_reason" simple string true if a previous effect set a reason for why an effect could not be completed.

Environment

Condition Type Description
"days_since_cataclysm" int or variable object true if at least days_since_cataclysm days have passed since the Cataclysm.
"is_season" string or variable object true if the current season matches is_season, which must be one of “spring", "summer", "autumn", or "winter".
"is_day" simple string true if it is currently daytime (sun is at or above the civil dawn point)
"u_is_outside"
"npc_is_outside"		 simple string true if you or the NPC is on a tile without a roof.
"u_is_underwater"
"npc_is_underwater"		 simple string true if you or the NPC is underwater.
"one_in_chance" int or variable object true if a one in one_in_chance random chance occurs.
"x_in_y_chance" object true if a x in y random chance occurs. x and y are either ints or variable object.
"is_weather" string or variable object true if current weather is "is_weather".

Meta

Condition Type Description
"mod_is_loaded" string or variable object true if the mod with the given ID is loaded.
"get_condition" string or variable object Runs the condition stored in the variable get_condition for the current dialogue.
"get_game_option" string or variable object gets the true or false game option for the provided string.

Sample responses with conditions and effects 

{
  "text": "Understood.  I'll get those antibiotics.",
  "topic": "TALK_NONE",
  "condition": { "npc_has_effect": "infected" }
},
{
  "text": "I'm sorry for offending you.  I predict you will feel better in exactly one hour.",
  "topic": "TALK_NONE",
  "effect": { "npc_add_effect": "deeply_offended", "duration": 600 }
},
{
  "text": "Nice to meet you too.",
  "topic": "TALK_NONE",
  "effect": { "u_add_effect": "has_met_example_NPC", "duration": "PERMANENT" }
},
{
  "text": "Nice to meet you too.",
  "topic": "TALK_NONE",
  "condition": {
    "not": {
      "npc_has_var": "has_met_PC", "type": "general", "context": "examples", "value": "yes"
    }
  },
  "effect": {
    "npc_add_var": "has_met_PC", "type": "general", "context": "examples", "value": "yes"
  }
},
{
  "text": "[INT 11] I'm sure I can organize salvage operations to increase the bounty scavengers bring in!",
  "topic": "TALK_EVAC_MERCHANT_NO",
  "condition": { "u_has_intelligence": 11 }
},
{
  "text": "[STR 11] I punch things in face real good!",
  "topic": "TALK_EVAC_MERCHANT_NO",
  "condition": { "and": [ { "not": { "u_has_intelligence": 7 } }, { "u_has_strength": 11 } ] }
},
{ "text": "Maybe later.", "topic": "TALK_RANCH_WOODCUTTER", "condition": "npc_available" },
{
  "text": "[$8] I'll take a beer",
  "topic": "TALK_DONE",
  "condition": { "u_has_cash": 800 },
  "effect": { "u_buy_item": "beer", "container": "bottle_glass", "count": 2, "cost": 800 }
},
{
  "text": "Okay.  Lead the way.",
  "topic": "TALK_DONE",
  "condition": { "not": "at_safe_space" },
  "effect": "lead_to_safety"
},
{
  "text": "About one of those missions...",
  "topic": "TALK_MISSION_LIST_ASSIGNED",
  "condition": { "and": [ "has_many_assigned_missions", { "u_is_wearing": "badge_marshal" } ] }
},
{
  "text": "[MISSION] The captain sent me to get a frequency list from you.",
  "topic": "TALK_OLD_GUARD_NEC_COMMO_FREQ",
  "condition": {
    "and": [
      { "u_is_wearing": "badge_marshal" },
      { "u_has_mission": "MISSION_OLD_GUARD_NEC_1" },
      { "not": { "u_has_effect": "has_og_comm_freq" } }
    ]
  }
},
{
  "text": "I killed them.  All of them.",
  "topic": "TALK_MISSION_SUCCESS",
  "condition": {
    "and": [ { "or": [ { "mission_goal": "KILL_MONSTER_SPEC" }, { "mission_goal": "KILL_MONSTER_TYPE" } ] }, "mission_complete" ]
  },
  "switch": true
},
{
  "text": "Glad to help.  I need no payment.",
  "topic": "TALK_NONE",
  "effect": "clear_mission",
  "mission_opinion": { "trust": 4, "value": 3 },
  "opinion": { "fear": -1, "anger": -1 }
},
{
  "text": "Maybe you can teach me something as payment?",
  "topic": "TALK_TRAIN",
  "condition": { "or": [ "npc_train_skills", "npc_train_styles" ] },
  "effect": "mission_reward"
},
{
  "truefalsetext": {
    "true": "I killed him.",
    "false": "I killed it.",
    "condition": { "mission_goal": "ASSASSINATE" }
  },
  "condition": {
    "and": [
      "mission_incomplete",
      {
        "or": [
          { "mission_goal": "ASSASSINATE" },
          { "mission_goal": "KILL_MONSTER" },
          { "mission_goal": "KILL_MONSTER_SPEC" },
          { "mission_goal": "KILL_MONSTER_TYPE" }
        ]
      }
    ]
  },
  "trial": { "type": "LIE", "difficulty": 10, "mod": [ [ "TRUST", 3 ] ] },
  "success": { "topic": "TALK_NONE" },
  "failure": { "topic": "TALK_MISSION_FAILURE" }
},
{
  "text": "Didn't you say you knew where the Vault was?",
  "topic": "TALK_VAULT_INFO",
  "condition": { "not": { "u_has_var": "asked_about_vault", "value": "yes", "type": "sentinel", "context": "old_guard_rep" } },
  "effect": [
    { "u_add_var": "asked_about_vault", "value": "yes", "type": "sentinel", "context": "old_guard" },
    { "mapgen_update": "hulk_hairstyling", "om_terrain": "necropolis_a_13", "om_special": "Necropolis", "om_terrain_replace": "field", "z": 0 }
  ]
},
{
  "text": "Why do zombies keep attacking every time I talk to you?",
  "topic": "TALK_RUN_AWAY_MORE_ZOMBIES",
  "condition": { "u_has_var": "even_more_zombies", "value": "yes", "type": "trigger", "context": "learning_experience" },
  "effect": [
    { "mapgen_update": [ "even_more_zombies", "more zombies" ], "origin_npc": true },
    { "mapgen_update": "more zombies", "origin_npc": true, "offset_x": 1 },
    { "mapgen_update": "more zombies", "origin_npc": true, "offset_x": -1 },
    { "mapgen_update": "more zombies", "origin_npc": true, "offset_y": 1 },
    { "mapgen_update": "more zombies", "origin_npc": true, "offset_y": -1 }
  ]
}

Utility Structures

Variable Object

variable_object: This is either an object, an arithmetic/math expression or array describing a variable name. It can either describe a double, a time duration or a string. If it is an array it must have 2 values the first of which will be a minimum and the second will be a maximum, the value will be randomly between the two. If it is a double default is a double which will be the value returned if the variable is not defined. If is it a duration then default can be either an int or a string describing a time span. u_val, npc_val, context_val, var_val or global_val can be the used for the variable name element. If u_val is used it describes a variable on player u, if npc_val is used it describes a variable on player npc, if context_val is used it describes a variable on the current dialogue context, var_val tries to resolve a variable stored in a context_val (more explanation bellow), if global_val is used it describes a global variable. If this is a duration infinite will be accepted to be a virtually infinite value(it is actually more than a year, if longer is needed a code change to make this a flag or something will be needed).

Example:

"effect": [ { "u_mod_focus": { "u_val":"test", "default": 1 } },
  { "u_mod_focus": [ 0, { "u_val":"test", "default": 1 } ] }
  { "u_add_morale": "morale_honey","bonus": -20,"max_bonus": -60, "decay_start": 1 },
  "duration": { "global_val": "test2", "default": "2 minutes" },
  {
    "u_spawn_monster": "mon_absence",
    "real_count": { "arithmetic": [ { "arithmetic": [ { "const":1 }, "+", { "const": 1 } ] }, "+", { "const": 1 } ] }
  }
]

var_val

var_val is a unique variable object in the fact that it attempts to resolve the variable stored inside a context variable. So if you had

Name Type Value
ref context_val key1
ref2 context_val u_key2
key1 global_val SOME TEXT
key2 u_val SOME OTHER TEXT

If you access “ref” as a context val it will have the value of “key1”, if you access it as a var_val it will have a value of “SOME TEXT”. If you access “ref2” as a context val it will have the value of “u_key2”, if you access it as a var_val it will have a value of “SOME OTHER TEXT”.

The values for var_val use the same syntax for scope that math variables do.

Mutators

mutators: take in an amount of data and provide you with a relevant string. This can be used to get information about items, monsters, etc. from the id, or other data. Mutators can be used anywhere that a string variable object can be used. Mutators take the form:

{ "mutator": "MUTATOR_NAME", "REQUIRED_KEY1": "REQUIRED_VALUE1", ..., "REQUIRED_KEYn": "REQUIRED_VALUEn" }

List Of Mutators

Mutator Name Required Keys Description
"mon_faction" mtype_id: String or variable object. Returns the faction of the monster with mtype_id.
"game_option" option: String or variable object. Returns the value of the option as a string, for numerical options you should instead use the math function.
"ma_technique_name" matec_id: String or variable object. Returns the name of the martial arts tech with ID matec_id
"ma_technique_description" matec_id: String or variable object. Returns the description of the martial arts tech with ID matec_id
"valid_technique” blacklist: array of String or variable object.
crit: bool
dodge_counter: bool
block_counter: bool		 Returns a random valid technique for the alpha talker to use against the beta talker with the provided specifications.
"loc_relative_u" target: String or variable object. target should be a string like “(x,y,z)” where x,y,z are coordinates relative to the player. Returns the abs_ms coordinates as a string (ready to store as a location variable), in the form “(x,y,z)” of the provided point relative to the player. So "target":"(0,1,0)" would return the point south of the player.
"topic_item"   Returns current topic_item as a string. See Repeat Responses

Compare Numbers and Arithmetics

arithmetic and compare_num are deprecated in the long term. See Math for the replacement.

"compare_num" can be used to compare two values to each other, while "arithmetic" can be used to take up to two values, perform arithmetic on them, and then save them in a third value. The syntax is as follows.

{
  "text": "If player strength is more than or equal to 5, sets time since cataclysm to the player's focus times the player's maximum mana with at maximum a value of 15.",
  "topic": "TALK_DONE",
  "condition": { "compare_num": [ { "u_val": "strength" }, ">=", { "const": 5 } ] },
  "effect": { "arithmetic": [ { "time_since_cataclysm": "turns" }, "=", { "u_val": "focus" }, "*", { "u_val": "mana_max" } ], "max":15 }
},

min and max are optional double or variable_object values. If supplied they will limit the result, it will be no lower than min and no higher than max. min_time and max_time work the same way but will parse times written as a string i.e. “10 hours”. "compare_num" supports the following operators: "==", "=" (Both are treated the same, as a compare), "!=", "<=", ">=", "<", and ">".

"arithmetic" supports the following operators: "*"(multiplication), "/"(division), "+"(addition), "-"(subtraction), "%"(modulus), "^"(power) and the following results "=", "*=", "/=", "+=", "-=", "%=", "++", and "--"

To get player character properties, use "u_val". To get NPC properties, use same syntax but "npc_val" instead. For vars only global_val is also allowed.

List of values that can be read and/or written to

Example Description
"const": 5 A constant value, in this case 5. Can be read but not written to.
"time": "5 days" A constant time value. Will be converted to turns. Can be read but not written to.
"time_since_cataclysm": "turns" Time since the start of the cataclysm in turns. Can instead take other time units such as minutes, hours, days, weeks, seasons, and years.
"time_until_eoc": "eoc_id" Time from now until the next scheduled run of eoc_id. Optional member unit can specify a member to use instead of turns such as minutes, hours, days, weeks, seasons, and years.
"rand": 20 A random value between 0 and a given value, in this case 20. Can be read but not written to.
"faction_trust": "free_merchants" The trust the faction has for the player (see FACTIONS.md) for details.
"faction_like": "free_merchants" How much the faction likes the player (see FACTIONS.md) for details.
"faction_respect": "free_merchants" How much the faction respects the player(see FACTIONS.md) for details.
"u_val": "strength" Player character’s strength. Can be read but not written to. Replace "strength" with "dexterity", "intelligence", or "perception" to get such values.
"u_val": "strength_base" Player character’s strength. Replace "strength_base" with "dexterity_base", "intelligence_base", or "perception_base" to get such values.
"u_val": "strength_bonus" Player character’s current strength bonus. Replace "strength_bonus" with "dexterity_bonus", "intelligence_bonus", or "perception_bonus" to get such values.
"u_val": "var" Custom variable. "var_name", "type", and "context" must also be specified. If global_val is used then a global variable will be used. If default is given as either an int or a variable_object then that value will be used if the variable is empty. If default_time is the same thing will happen, but it will be parsed as a time string aka “10 hours”. Otherwise 0 will be used if the variable is empty.
"u_val": "time_since_var" Time since a custom variable was set. Unit used is turns. "var_name", "type", and "context" must also be specified.
"u_val": "allies" Number of allies the character has. Only supported for the player character. Can be read but not written to.
"u_val": "cash" Amount of money the character has. Only supported for the player character. Can be read but not written to.
"u_val": "owed" Owed money to the NPC you’re talking to.
"u_val": "sold" Amount sold to the NPC you’re talking to.
"u_val": "hp" Amount of hp. If bodypart is provided it will be for that part otherwise it will be the sum of all parts.
"u_val": "warmth" Amount of warmth in a given bodypart. bodypart is the id of the part to use.
"u_val": "effect_intensity" Intensity of an effect. effect is the id of the effect to test and bodypart is optionally the body part to look at. If the effect is not present a -1 is returned.
"u_val": "dodge" Current effective dodge of the character.
"u_val": "pos_x" Player character x coordinate. “pos_y” and “pos_z” also works as expected.
"u_val": "power" Bionic power in millijoule.
"u_val": "power_max" Max bionic power in millijoule. Can be read but not written to.
"u_val": "power_percentage" Percentage of max bionic power. Should be a number between 0 to 100.
"u_val": "morale" The current morale. Can be read but not written to for players and for monsters can be read and written to.
"u_val": "mana" Current mana.
"u_val": "mana_max" Max mana. Can be read but not written to.
"u_val": "hunger" Current perceived hunger. Can be read but not written to.
"u_val": "thirst" Current thirst.
"u_val": "instant_thirst" Current thirst minus water in the stomach that hasn’t been absorbed by the body yet.
"u_val": "stored_kcal" Stored kcal in the character’s body. 55’000 is considered healthy.
"u_val": "stored_kcal_percentage" a value of 100 represents 55’000 kcal, which is considered healthy.
"u_val": "item_count" Number of a given item in the character’s inventory. "item" must also be specified. Can be read but not written to.
"u_val": "charge_count" Number of charges of a given item in the character’s inventory. "item" must also be specified. Can be read but not written to.
"u_val": "exp" Total experience earned.
"u_val": "addiction_intensity", "addiction": "caffeine" Current intensity of the given addiction. Allows for an optional field "mod" which accepts an integer to multiply against the current intensity.
"u_val": "addiction_turns", "addiction": "caffeine" Current duration left (in turns) for the given addiction.
"u_val": "stim" Current stim level.
"u_val": "pkill" Current painkiller level.
"u_val": "rad" Current radiation level.
"u_val": "focus" Current focus level.
"u_val": "activity_level" Current activity level index as a floored integer, from 0-5. Roughly: 0.45 = SLEEP_EXERCISE (floored and returns 0), 0.5 = NO_EXERCISE(floored and returns 0), 1 = LIGHT_EXERCISE, 2 = MODERATE_EXERCISE, 3 = BRISK_EXERCISE, 4 = ACTIVE_EXERCISE, 5 = EXTRA_EXERCISE.
"u_val": "fatigue" Current fatigue level.
"u_val": "stamina" Current stamina level.
"u_val": "health" Current health level.
"u_val": "sleep_deprivation" Current sleep deprivation level.
"u_val": "anger" Current anger level, only works for monsters.
"u_val": "friendly" Current friendly level, only works for monsters.
"u_val": "npc_anger" Current anger level the npc has.
"u_val": "npc_trust" Current trust the npc has for you.
"u_val": "npc_fear" Current fear of the npc.
"u_val": "npc_value" Current value npc places on you.
"u_val": "vitamin" Current vitamin level. name must also be specified which is the vitamins id.
"u_val": "fine_detail_vision_mod" Returned values range from 1.0 (unimpeded vision) to 11.0 (totally blind).
"u_val": "age" Current age in years.
"u_val": "body_temp" Current body temperature.
"u_val": "body_temp_delta" Difference in temperature between the hottest/coldest part and what feels like the hottest/coldest part.
"u_val": "bmi_permil" Current BMI per mille (Body Mass Index x 1000)
"u_val": "height" Current height in cm. When setting there is a range for your character size category. Setting it too high or low will use the limit instead. For tiny its 58, and 87. For small its 88 and 144. For medium its 145 and 200. For large its 201 and 250. For huge its 251 and 320.
"u_val": "size" Size category from 1 (tiny) to 5 (huge). Read-only.
"u_val": "grab_strength" Grab strength as defined in the monster definition. Read-only, returns false on characters.
"u_val": "field_strength" Strength of a field on the tile the player or NPC is standing on. field must be specified. read only.
"u_val": "spell_level" Level of a given spell. -1 means the spell is not known when read and that the spell should be forgotten if written. Optional params: school gives the highest level of spells known of that school (read only), spell reads or writes the level of the spell with matching spell id. If no parameter is provided, you will get the highest spell level of the spells you know (read only).
"u_val": "spell_exp" Experience for a given spell. -1 means the spell is not known when read and that the spell should be forgotten if written. Required param: spell is the id of the spell in question.
"u_val": "spell_level_adjustment" Temporary caster level adjustment. Only useable by EoCs that trigger on the event opens_spellbook. Old values will be reset to 0 before the event triggers. To avoid overwriting values from other EoCs, it is recommended to adjust the values here with += or -= instead of setting it to an absolute value. When an NPC consider what spell to cast they will be considered the primary talker, so their values are manipulated with u_val the same way the player’s values are. Optional params: school makes it only apply to a given school. spell makes it only apply to a given spell.
"u_val": "spell_count" Number of spells that this character knows. Optional params: school returns only the number of spells of the specified spell class that the character knows. Read-only.
"u_val": "proficiency" Deals with a proficiency. Required params: proficiency_id is the id of the proficiency dealt with. format determines how the proficiency will be interacted with. "format": <int> will read or write how much you have trained a proficiency out of . So for example, if you write a 5 to a proficiency using `"format": 10`, you will set the proficiency to be trained to 50%. `"format": "percent"` reads or writes how many percent done the learning is. `"format": "permille"` does likewise for permille. `"format": "total_time_required"` gives you total time required to train a given proficiency (read only). `"format": "time_spent"` deals with total time spent. `"format": "time_left"` sets the remaining time instead. For most formats possible, If the resulting time is set to equal or more than the time required to learn the proficiency, you learn it. If you read it and it gives back the total time required, it means it is learnt. Setting the total time practiced to a negative value completely removes the proficiency from your known and practiced proficiencies. If you try to read time spent on a proficiency that is not in your proficiency list, you will get back 0 seconds.
"u_val": "volume" Current volume in ml. read only. Cullently, doesn’t work for characters, but for monsters and items.
"u_val": "weight" Current weight in mg. read only.
"distance": [] Distance between two targets. Valid targets are: “u”,”npc” and an object with a variable name.

Example:<pre>“condition”: { “compare_num”: [
{ “distance”: [ “u”,{ “u_val”: “stuck”, “type”: “ps”, “context”: “teleport” } ] },
“>”, { “const”: 5 }
] }</pre>
"hour" Hours since midnight.
"moon" Phase of the moon.<pre>MOON_NEW =0,
WAXING_CRESCENT =1,
HALF_MOON_WAXING =2,
WAXING_GIBBOUS =3,
FULL =4,
WANING_GIBBOUS =5,
HALF_MOON_WANING =6,
WANING_CRESCENT =7</pre>
"mod_load_order" This should be a string with the name of the mod. It will return the order it was loaded in or -1 if its not loaded.
"arithmetic" An arithmetic expression with no result.

Example:<pre>“real_count”: { “arithmetic”: [
{ “arithmetic”: [ { “const”:1 }, “+”, { “const”: 1 } ] },
“+”, { “const”: 1 }
] },</pre>
"math" An array math object.

Math

A math object lets you evaluate math expressions and assign them to dialogue variables or compare them for conditions. It takes the form

{ "math": [ "2 + 2 - 3", "==", "1" ] }

or idiomatically

{ "math": [ "lhs", "operator", "rhs" ] }

It takes an array as a parameter with 1, 2, or 3 strings:

One string = return value 

{ "math": [ "lhs" ] }

Example:

{ "value": "LUMINATION", "add": { "math": [ "u_val('morale') * 3 - rng(0, 100)" ] } }

The expression in lhs is evaluated and passed on to the parent object.

Two strings = unary operation 

{ "math": [ "lhs", "operator" ] }

Example:

"effect": [ { "math": [ "math_test", "++" ] } ]

lhs must be an assignment target. operator can be ++ or -- to increment or decrement, respectively.

Three strings = assignment or comparison 

{ "math": [ "lhs", "operator", "rhs" ] }

If operator is =, +=, -=, *=, /=, or %= the operation is an assignment:

"effect": { "math": [ "u_blorg", "=", "rng( 0, 2 ) + u_val('spell_level', 'spell: test_spell_pew') / 2" ] }

lhs must be an assignment target. rhs is evaluated and stored in the assignment target from lhs.

If operator is ==, !=, >=, <=, >, or <, the operation is a comparison:

"condition": { "math": [ "u_val('stamina') * 2", ">=", "5000 + rand( 300 )" ] },

lhs and rhs are evaluated independently and the result of operator is passed on to the parent object.

Variables

Tokens that aren’t numbers, constants, functions, or mathematical symbols are treated as dialogue variables. They are scoped by their name so myvar is a variable in the global scope, u_myvar is scoped on the alpha talker, n_myvar is scoped on the beta talker, and _var is a context variable, v_var is a var_val.

Examples:

    "//0": "return value of global var blorgy_counter",
    { "math": [ "blorgy_counter" ] },
    "//1": "result, x, and y are global variables",
    { "math": [ "result", "=", "x + y" ] },
    "//2": "u_z is the variable z on the alpha talker (avatar)",
    { "math": [ "result", "=", "( x + y ) * u_z" ] },
    "//3": "n_crazyness is the variable crazyness on the beta talker (npc)",
    { "math": [ "n_crazyness * 2", ">=", "( x + y ) * u_z" ] },

Constants

The tokens π (and pi alias ) and e are recognized as mathematical constants and get replaced by their nominal values.

Math functions

Common math functions are supported:

abs(), sqrt(), log(), floor(), trunc(), ceil(), round(), sin(), cos(), and tan() take one argument, for example sin( 2 * pi + 1 ).

floor( x ) returns the smallest integer not greater than x, for example floor( 1.5 ) is 1, floor( -1.5 ) is -2

ceil( x ) returns the smallest integer not less than x, for example ceil( 1.5 ) is 2, ceil( -1.5 ) is -1

trunc( x ) returns the nearest integer closer to zero, for example trunc( 1.5 ) is 1, trunc( -1.5 ) is -1

clamp( x, lo, hi ) clamps x between lo and hi. hi must be greater than lo

max() and min() take any number of arguments, for example max( 1, 2, 3 ).

rand() takes one argument x and returns a random integer between 0 and x, for example rand(100).

rng() takes two arguments a and b and returns a random floating point value between a and b, for example rng(1.5, 2).

Function composition is also supported, for example sin( rng(0, max( 0.5, u_sin_var ) ) )

Ternary and inline boolean operators

Inline comparison operators evaluate as 1 for true and 0 for false.

Ternary operators take the form condition ? true_value : false_value. They are right-associative so a chained ternary like a ? b : c ? d :e is parsed as a ? b : (c ? d : e).

Examples:

    "//0": "returns 5 if u_blorg is greater than 4, otherwise 0",
    { "math": [ "( u_blorg > 4 ) * 5" ] },
    "//1": "returns rng( 0.5, 5 ) if u_blorg is greater than 5, otherwise rand(100)"
    { "math": [ "u_blorg > 5 ? rng( 0.5, 5 ) : rand(100)" ] },

Dialogue functions

Dialogue functions return or manipulate game values. They are scoped just like variables.

These functions support keyword-value pairs as optional arguments (kwargs) of the form 'keyword': argument.

This section is a work in progress as functions are ported from arithmetic to math.

function arguments are doubles (or sub-expressions), strings, or variables
some functions support array arguments or kwargs, denoted with square brackets [] (ex: [d/v])

Function Eval Assign Scopes Description
armor(s/v,s/v) ✅ ❌ u, n Return the numerical value for a characters armor on a body part, for a damage type.
Arguments are damagetype ID, bodypart ID.

Example:
"condition": { "math": [ "u_armor('bash', 'torso')", ">=", "5"] }
attack_speed() ✅ ❌ u, n Return the characters current adjusted attack speed with their current weapon.

Example:
"condition": { "math": [ "u_attack_speed()", ">=", "10"] }
addiction_intensity(s/v) ✅ ❌ u, n Return the characters current intensity of the given addiction.
Argument is addiction type ID.

Example:
"condition": { "math": [ "u_addiction_intensity('caffeine')", ">=", "1"] }
addiction_turns(s/v) ✅ ✅ u, n Return the characters current duration left (in turns) for the given addiction.
Argument is addiction type ID.

Example:
"condition": { "math": [ "u_addiction_turns('caffeine')", ">=", "3600"] }
charge_count(s/v) ✅ ❌ u, n Return the charges of a given item in the character’s inventory.
Argument is item ID.

Example:
"condition": { "math": [ "u_charge_count('light_plus_battery_cell')", ">=", "100"] }
coverage(s/v) ✅ ❌ u, n Return the characters total coverage of a body part.
Argument is bodypart ID.
For items, returns typical coverage of the item.

Example:
"condition": { "math": [ "u_coverage('torso')", ">", "0"] }
distance(s/v,s/v) ✅ ❌ g Return distance between two targets.
Arguments are location variables or special strings (u, npc). u means your location. npc means NPC’s location.

Example:
"condition": { "math": [ "distance('u', loc)", "<=", "50"] }
effect_intensity(s/v) ✅ ❌ u, n Return the characters intensity of effect.
Argument is effect ID.

Optional kwargs:
bodypart: s/v - Specify the bodypart to get/set intensity of effect.

Example:
"condition": { "math": [ "u_effect_intensity('bite', 'bodypart': 'torso')", ">", "1"] }
encumbrance(s/v) ✅ ❌ u, n Return the characters total encumbrance of a body part.
Argument is bodypart ID.
For items, returns typical encumbrance of the item.

Example:
"condition": { "math": [ "u_encumbrance('torso')", ">", "0"] }
field_strength(s/v) ✅ ❌ u, n, global Return the strength of a field on the tile.
Argument is field ID.

Optional kwargs:
location: v - center search on this location

The location kwarg is mandatory in the global scope.

Examples:
"condition": { "math": [ "u_field_strength('fd_blood')", ">", "5" ] }

"condition": { "math": [ "field_strength('fd_blood_insect', 'location': u_search_loc)", ">", "5" ] }
has_trait(s/v) ✅ ❌ u, n Check whether the actor has a trait. Meant to be used as condition for ternaries. Arguemnt is trait ID.

Example:
"condition": { "math": [ "u_blorg", "=", "u_has_trait('FEEBLE') ? 100 : 15" ] }
has_proficiency(s/v) ✅ ❌ u, n Check whether the actor has a proficiency. Meant to be used as condition for ternaries. Arguemnt is proficiency ID.

Example:
"condition": { "math": [ "u_blorg", "=", "u_has_proficiency('prof_intro_biology') ? 100 : 15" ] }
hp(s/v) ✅ ✅ u, n Return or set the characters hp. Argument is bodypart ID. For special value ALL, get hp sum of all bodyparts or set hp of all bodyparts.

For items, returns current amount of damage required to destroy item.

Example:
"condition": { "math": [ "hp('torso')", ">", "100"] }
hp_max(s/v) ✅ ❌ u, n Return the characters max amount of hp on a body part.
Argument is bodypart ID.
For items, returns max amount of damage required to destroy item.

Example:
"condition": { "math": [ "u_hp_max('torso')", ">=", "100"] }
game_option(s/v) ✅ ❌ N/A
(global)		 Return the numerical value of a game option

Example:
"condition": { "math": [ "game_option('NPC_SPAWNTIME')", ">=", "5"] }
item_count(s/v) ✅ ❌ u, n Return the number of a given item in the character’s inventory.
Argument is item ID.

Example:
"condition": { "math": [ "u_item_count('backpack')", ">=", "1"] }
monsters_nearby(s/v…) ✅ ❌ u, n, global Return the number of nearby monsters. Takes any number of string or variable positional parameters as monster IDs.

Optional kwargs:
radius: d/v - limit to radius (rl_dist)
location: v - center search on this location

The location kwarg is mandatory in the global scope.

Examples:
"condition": { "math": [ "u_monsters_nearby('radius': u_search_radius * 3)", ">", "5" ] }

"condition": { "math": [ "monsters_nearby('mon_void_maw', 'mon_void_limb', mon_fotm_var, 'radius': u_search_radius * 3, 'location': u_search_loc)", ">", "5" ] }
num_input(s/v,d/v) ✅ ❌ N/A
(global)		 Prompt the player for a number.
Arguments are Prompt text, Default Value:
"math": [ "u_value_to_set", "=", "num_input('Playstyle Perks Cost?', 4)" ]
pain() ✅ ✅ u, n Return or set pain
Example:
{ "math": [ "n_pain()", "=", "u_pain() + 9000" ] }
proficiency(s/v) ✅ ✅ u, n Return or set proficiency
Argument is proficiency ID.

Optional kwargs:
format: s - percent return or set how many percent done the learning is. permille does likewise for permille. time_spent return or set total time spent. time_left return or set the remaining time. total_time_required return total time required to train a given proficiency (read only).

Example:
{ "math": [ "u_proficiency('prof_intro_chemistry', 'format': 'percent')", "=", "50" ] }
school_level(s/v) ✅ ❌ u, n Return the highest level of spells known of that school.
Argument is school ID.

Example:
"condition": { "math": [ "u_school_level('MAGUS')", ">=", "3"] }
school_level_adjustment(s/v) ✅ ✅ u, n Return or set temporary caster level adjustment. Only useable by EoCs that trigger on the event opens_spellbook. Old values will be reset to 0 before the event triggers. To avoid overwriting values from other EoCs, it is recommended to adjust the values here with += or -= instead of setting it to an absolute value.
Argument is school ID.

Example:
{ "math": [ "u_school_level_adjustment('MAGUS')", "+=", "3"] }
skill(s/v) ✅ ✅ u, n Return or set skill level

Example:
"condition": { "math": [ "u_skill('driving')", ">=", "5"] }
"condition": { "math": [ "u_skill(someskill)", ">=", "5"] }
skill_exp(s/v) ✅ ✅ u, n Return or set skill experience
Argument is skill ID.

Optional kwargs:
format: s/v - must be percentage or raw.

Example:
{ "math": [ "u_skill_exp('driving', 'format': crt_format)", "+=", "10"] }
spell_exp(s/v) ✅ ✅ u, n Return or set spell xp
Example:
"condition": { "math": [ "u_spell_exp('SPELL_ID')", ">=", "5"] }
spell_count() ✅ ❌ u, n Return number of spells the character knows.

Optional kwargs:
school: s/v - return number of spells known of that school.

Example:
"condition": { "math": [ "u_spell_count('school': 'MAGUS')", ">=", "10"] }
spell_level(s/v) ✅ ✅ u, n Return or set level of a given spell. -1 means the spell is not known when read and that the spell should be forgotten if written.
Argument is spell ID. If "null" is given, return the highest level of spells the character knows (read only).
Example:
"condition": { "math": [ "u_spell_level('SPELL_ID')", "==", "-1"] }
spell_level_adjustment(s/v) ✅ ✅ u, n Return or set temporary caster level adjustment. Only useable by EoCs that trigger on the event opens_spellbook. Old values will be reset to 0 before the event triggers. To avoid overwriting values from other EoCs, it is recommended to adjust the values here with += or -= instead of setting it to an absolute value.
Argument is spell ID. If "null" is given, adjust all spell level.

Example:
{ "math": [ "u_spell_level_adjustment('SPELL_ID')", "+=", "3"] }
spellcasting_adjustment(s/v) ❌ ✅ u Temporary alters a property of spellcasting. Only useable by EoCs that trigger on the event opens_spellbook. Old values will be reset to default values before the event triggers. Multipliers have a default value of 1, while adjustments have a default value of 0. Assignment functions as an adjustment to the default value by the value assigned. So a = functions as you would expect a += to function. Reading these values are not possible, and therefore using += is not possible.

Possible argument values:
caster_level - Adjustment - alters the caster level of the given spell(s). Works much like spell_level_adjustment, but will not be readable by math functions.
casting_time - Multiplier - alters the casting time of the given spell(s).
cost - Multiplier - alters the cost of the given spells. Note that this does not change what items may be consumed by the spell(s).
aoe - Multiplier - alters the area of effect of the spell(s).
range - Multiplier - alters the range of the spell(s).
duration - Multiplier - alters the duration of the spell(s).
difficulty - Adjustment - alters the difficulty of the spell(s), thus altering the probability of failing spellcasting.
somatic_difficulty - Multiplier - alters how much encumbrance affects spellcasting time and difficulty. If set to 0, it will also remove the need to have your hands free while casting. Note that as a multiplier, it starts of at 1, and setting the value actually adjusts it. So setting the valute to -1 will result in a final value of 0. Alternatively, setting it to -0,5 twice would also do the trick.
sound - Multiplier - alters the loudness and how much mouth encumbrance affects the spell(s).
concentration - Multiplier - alters how much focus alters the difficulty of the spell(s).

Optional kwargs:
flag_blacklist: s/v and
flag_whitelist: s/v - Only applies the modifier to spells that matches the blacklist and/or whitelist
mod: s/v, school: s/v, spell: s/v - Only one of these can be applied. Limits what spells will be affected. If none are specified, the modification will apply to all spells (whitelist and blacklist still applies separately).

Example:
{ "math": [ "u_spellcasting_adjustment('casting_time', 'mod': 'magiclysm', 'flag_blacklist': 'CONSUMES_RUNES' )", "=", "-0.95" ] }
vitamin(s/v) ✅ ✅ u, n Return or set the characters vitamin level.
Argument is vitamin ID.

Example:
{ "math": [ "u_vitamin('mutagen')", "=", "0" ] }
warmth(s/v) ✅ ❌ u, n Return the characters warmth on a body part.
Argument is bodypart ID.

Example:
The value displayed in-game is calculated as follows.
"{ "math": [ "u_warmth_in_game", "=", "(u_warmth('torso') / 100) * 2 - 100"] }
weather(s) ✅ ✅ N/A
(global)		 Return or set a weather aspect

Aspect must be one of:
temperature (in Kelvin),
humidity (as percentage),
pressure (in millibar),
windpower (in mph).
precipitation (in mm / h) either 0.5 (very_light ), 1.5 (light), or 3 (heavy). Read only.

Temperature conversion functions are available: celsius(), fahrenheit(), from_celsius(), and from_fahrenheit().

Examples:
{ "math": [ "weather('temperature')", "<", "from_fahrenheit( 33 )" ] }
{ "math": [ "fahrenheit( weather('temperature') )", "==", "21" ] }
u_val shim

There is a val() shim available that can cover the missing arithmetic functions from u_val and npc_val:

{ "math": [ "u_val('stamina')" ] }

where

"u_val('stamina')"

is equivalent to

{ "u_val": "stamina" }

Keyword-value pairs are also supported:

"n_val('spell_level', 'spell: test_spell_pew')"

is equivalent to

{ "npc_val": "spell_level", "spell": "test_spell_pew" }

Most of the values from arithmetic can be used with the shim, both for reading and writing.

More examples:

    { "math": [ "u_val('age')" ] },
    { "math": [ "u_val('time: 1 d')" ] },
    { "math": [ "u_val('proficiency', 'proficiency_id: prof_test', 'format: percent')" ] }

Math functions defined in JSON

Math functions can be defined in JSON like this

  {
    "type": "jmath_function",
    "id": "my_math_function",
    "num_args": 2,
    "return": "_0 * 2 + rand(_1)"
  },

where _0, _1, etc. are positional parameters.

These functions can then be used like regular math functions, for example:

  {
    "type": "effect_on_condition",
    "id": "EOC_do_something",
    "effect": [ { "math": [ "secret_value", "=", "my_math_function( u_pain(), 500)" ] } ]
  },

so _0 takes the value of u_pain() and _1 takes the value 500 inside my_math_function()

Function composition is also supported and the functions can be defined and used in any order.

Assignment target

An assignment target can be either a scoped variable name or a scoped dialogue function.