Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.
Python agent-contextagent-memoryagent-pluginsagentic-ragcontext-databasedsh-pluginknowledge-basememorymemory-managementself-evolvingvector-database
import { normalizeDirUri } from '#/lib/viking-uri'
import { KNOWN_VIKING_SCOPES } from '../-constants/retrieval'
import type { RetrievalScope } from '../-types/retrieval'
export function normalizeScopeInput(value: string): string | undefined {
const trimmed = value.trim()
if (!trimmed || trimmed === '/' || trimmed === 'viking://') {
return undefined
}
if (trimmed.startsWith('viking://')) {
return normalizeDirUri(trimmed)
}
const path = trimmed.replace(/^\/+/, '')
if (!path) {
return undefined
}
const [scope] = path.split('/')
const scopedPath = KNOWN_VIKING_SCOPES.has(scope) ? path : `resources/${path}`
return normalizeDirUri(`viking://${scopedPath}`)
}
export function resolveScopeTargetUri(
scope: RetrievalScope,
customPathInput: string,
): string | undefined {
if (scope === 'all') {
return undefined
}
if (scope === 'resources') {
return 'viking://resources/'
}
return normalizeScopeInput(customPathInput)
}