| Option |
Type |
Default |
Description |
maxBodyBytes |
number |
1_000_000 |
Max bytes read from message body |
forAIMaxChars |
number |
8_000 |
Max chars in content.forAI |
cleaner |
fn |
built-in |
Custom quote/signature stripper |
wrapper |
ForAIWrapper |
none |
Delimiter wrapper for forAI |
onAttachment |
fn |
none |
Callback per attachment during parse |
attachmentsInForAI |
boolean |
false |
Append att.extractedText to content.forAI (roadmap — attachment processors, §9.3–9.8; not in the current ParseOptions type) |
attachmentsForAIOptions |
AttachmentsForAIOptions |
defaults |
Controls per-attachment limits and wrapping (roadmap — attachment processors, §9.3–9.8; not in the current ParseOptions type) |
threadIdResolver |
fn |
AECS-1 §5 |
Custom threadId calculation |
specVersion |
string |
SDK default |
Stamp in processing.specVersion |
| Option |
Type |
Default |
Description |
maxMessages |
number |
all |
Limit to N most recent messages |
maxCharsPerMessage |
number |
2_000 |
Truncate each message |
wrapper |
ForAIWrapper |
none |
Per-message wrapper |
includeMetadata |
boolean |
true |
Prepend sender + date per message |
order |
"asc" | "desc" |
"asc" |
Chronological or reverse |
Status: Roadmap. This section specifies a planned module; it is not yet implemented in @mvrx/mail.
| Option |
Type |
Default |
Description |
model |
string |
provider default |
Override model for this call |
tone |
Tone |
"professional" |
Writing tone |
length |
Length |
"standard" |
Response length target |
format |
"text" | "html" |
"text" |
Output format |
language |
string |
auto-detect |
ISO 639-1 target language |
systemPrompt |
string |
SDK default |
Prepend to system instructions |
maxTokens |
number |
1024 |
Cap LLM response tokens |
includeAttachments |
boolean |
false |
Pass att.extractedText from all attachments as additional LLM context |