What the model wrote:
const damageEvent = table(
{ name: 'damageEvent', public: true, event: true },
const account = table(
{ name: 'account', public: true },
const customer = table(
{ name: 'customer', public: true },
const orderDetail = table(
{ name: 'orderDetail', public: true },
const player = table(
{ name: 'player', public: true },
const primitive = table(
{ name: 'primitive', public: true },
const profile = table(
{ name: 'profile', public: true },
What was expected:
const damageEvent = table({
name: 'damage_event',
public: true,
event: true,
}, {
const account = table({
name: 'account',
}, {
const customer = table({
name: 'customer',
}, {
const orderDetail = table({
name: 'order_detail',
}, {
const player = table({
name: 'player',
}, {
const primitive = table({
name: 'primitive',
}, {
const profile = table({
name: 'profile',
}, {
What the error says:
Mostly schema parity failures such as:
schema_parity: describe failed: ... Invalid token: InvalidSignature
Why this happened:
The model mixed naming/style conventions and repeatedly added public: true where the schema did not call for it. It also failed to normalize table names to the snake_case names used in the golden schemas. This looks like a generic schema-hallucination issue rather than a runtime API error.
Affected tasks:
t_024_event_table, t_031_unique_constraint, t_030_two_table_join, t_025_optional_fields, t_017_scheduled_columns, t_014_elementary_columns, t_015_product_type_columns
AI guidelines gap:
The cheat-sheet likely did not emphasize exact table metadata formatting, especially when to omit public: true and when table names must be snake_case rather than camelCase.
What the model wrote:
const drawing = table(
{ name: 'drawing', public: true },
{
const order = table(
{ name: 'order', public: true },
{
const result = table(
{ name: 'result', public: true },
{
const tickTimer = table({
name: 'tickTimer',
scheduled: (): any => tick,
}, {
const log = table(
{
name: 'log',
public: true,
indexes: [
{
accessor: 'byUserDay',
algorithm: 'btree',
columns: ['userId', 'day']
}
]
},
const account = table(
{
name: 'account',
public: true,
indexes: [
{
accessor: 'byName',
algorithm: 'btree',
columns: ['name']
}
]
},
const membership = table(
{
name: 'membership',
public: true,
indexes: [
import { schema, table, t, ScheduleAt } from 'spacetimedb/server';
const spacetimedb = schema({ drawing });
What was expected:
const drawing = table({
name: 'drawing',
}, {
const order = table({
name: 'order',
}, {
const result = table({
name: 'result',
}, {
const tickTimer = table(
{
name: 'tick_timer',
scheduled: (): any => tick,
},
const log = table({
name: 'log',
indexes: [{ accessor: 'byUserDay', algorithm: 'btree', columns: ['userId', 'day'] }],
}, {
const account = table({
name: 'account',
indexes: [{ accessor: 'byName', algorithm: 'btree', columns: ['name'] }],
}, {
const membership = table(
{
name: 'membership',
indexes: [
import { ScheduleAt } from 'spacetimedb';
import { table, schema, t } from 'spacetimedb/server';
const spacetimedb = schema({ drawing });
What the error says:
The failures were largely publish/schema parity errors, e.g.:
publish_error: spacetime publish (typescript) failed
schema_parity: describe failed: ... Invalid token: InvalidSignature
Why this happened:
The model likely generalized from a different SpacetimeDB API shape and inserted metadata fields in the wrong object, or in the wrong import location. It also used incorrect import paths and mixed server/runtime symbols.
Affected tasks:
t_028_cascade_delete, t_021_multi_column_index, t_018_constraints, t_019_many_to_many, t_017_scheduled_columns, t_030_two_table_join, t_012_spacetime_product_type, t_013_spacetime_sum_type, t_024_event_table
AI guidelines gap:
The guidelines apparently did not clearly specify the exact table({...}, {...}) signature and which metadata keys belong in the first argument only, especially for indexes, scheduled tables, events, and imports.