@zigai/pi-message-highlights
Pi package that highlights URLs and file paths in messages and the prompt editor.
Package details
Install @zigai/pi-message-highlights from npm and Pi will load the resources declared by the package manifest.
$ pi install npm:@zigai/pi-message-highlights- Package
@zigai/pi-message-highlights- Version
0.4.0- Published
- Jul 17, 2026
- Downloads
- 386/mo · 83/wk
- Author
- zigai
- License
- MIT
- Types
- extension
- Size
- 240.2 KB
- Dependencies
- 2 dependencies · 2 peers
Pi manifest JSON
{
"extensions": [
"./src/index.ts"
]
}Security note
Pi packages can execute code and influence agent behavior. Review the source before installing third-party packages.
README
@zigai/pi-message-highlights
Highlights URLs and file paths in Pi's interactive UI:
- URLs render in a configurable blue foreground by default.
- File paths render with Pi's accent/highlight color.
- Applies to assistant responses, past user messages, and the prompt editor.
Install
pi install npm:@zigai/pi-message-highlights
Configuration
Global settings are stored in ~/.pi/agent/extension-settings/pi-message-highlights.json.
| Option | Type | Default | Description |
|---|---|---|---|
urlColor |
accent | bashMode | border | borderAccent | borderMuted | customMessageLabel | customMessageText | dim | error | mdCode | mdCodeBlock | mdCodeBlockBorder | mdHeading | mdHr | mdLink | mdLinkUrl | mdListBullet | mdQuote | mdQuoteBorder | muted | success | syntaxComment | syntaxFunction | syntaxKeyword | syntaxNumber | syntaxOperator | syntaxPunctuation | syntaxString | syntaxType | syntaxVariable | text | thinkingHigh | thinkingLow | thinkingMedium | thinkingMinimal | thinkingOff | thinkingText | thinkingXhigh | toolDiffAdded | toolDiffContext | toolDiffRemoved | toolOutput | toolTitle | userMessageText | warning | integer | `` | string |
"#87d7ff" |
URL color as an ANSI-256 index, hex color, theme color name, or empty string to disable highlighting. |
{
"$schema": "./schemas/pi-message-highlights.schema.json",
"urlColor": "#87d7ff"
}
License
MIT