feat: port source-first ingestion to os repo

- make source the canonical field across os routes, tools, ui, and mcp
- align fresh-install schema, search, and embedding flows with source-first ingestion

Generated with Claude Code
This commit is contained in:
“BeeRad”
2026-03-20 12:36:26 +11:00
parent 41f8498d4a
commit 28e570696c
34 changed files with 588 additions and 521 deletions
+21 -17
View File
@@ -77,7 +77,8 @@ All data stays on this device.`;
// Tool schemas
const addNodeInputSchema = {
title: z.string().min(1).max(160).describe('Clear, descriptive title'),
content: z.string().max(20000).optional().describe('Node content/notes'),
content: z.string().max(20000).optional().describe('Legacy content field; mapped to source'),
source: z.string().max(50000).optional().describe('Full source text'),
link: z.string().url().optional().describe('Source URL'),
description: z.string().min(24).max(280).describe('REQUIRED. One-sentence summary: WHAT this is (explicit, concrete) + WHY it matters. No weak verbs (discusses, explores, examines). Example: "Podcast — Lex Fridman interviews Sam Altman on AGI timelines. First public comments since board drama."'),
dimensions: z.array(z.string()).min(1).max(5).describe('1-5 categories. Call queryDimensions first to use existing ones.'),
@@ -354,7 +355,7 @@ async function main() {
// Note: MCP schema uses "content" for external API compat; mapped to "notes" internally
inputSchema: addNodeInputSchema
},
async ({ title, content, link, description, dimensions, metadata, chunk }) => {
async ({ title, content, source, link, description, dimensions, metadata, chunk }) => {
const normalizedDimensions = sanitizeDimensions(dimensions);
if (normalizedDimensions.length === 0) {
throw new Error('At least one dimension is required.');
@@ -366,12 +367,11 @@ async function main() {
const node = nodeService.createNode({
title: title.trim(),
notes: content?.trim(),
source: source?.trim() || chunk?.trim() || content?.trim(),
link: link?.trim(),
description: description?.trim(),
dimensions: normalizedDimensions,
metadata: metadata || {},
chunk: chunk?.trim()
metadata: metadata || {}
});
const summary = `Created node #${node.id}: ${node.title} [${node.dimensions.join(', ')}]`;
@@ -425,7 +425,7 @@ async function main() {
WITH fts_matches AS (
SELECT rowid, rank FROM nodes_fts WHERE nodes_fts MATCH ? LIMIT 100
)
SELECT n.id, n.title, n.description, n.notes, n.link,
SELECT n.id, n.title, n.description, n.source, n.link,
n.created_at, n.updated_at, n.event_date,
COALESCE((SELECT JSON_GROUP_ARRAY(d.dimension)
FROM node_dimensions d WHERE d.node_id = n.id), '[]') as dimensions_json
@@ -446,7 +446,7 @@ async function main() {
WITH fts_matches AS (
SELECT rowid, rank FROM nodes_fts WHERE nodes_fts MATCH ? LIMIT ?
)
SELECT n.id, n.title, n.description, n.notes, n.link,
SELECT n.id, n.title, n.description, n.source, n.link,
n.created_at, n.updated_at, n.event_date,
COALESCE((SELECT JSON_GROUP_ARRAY(d.dimension)
FROM node_dimensions d WHERE d.node_id = n.id), '[]') as dimensions_json
@@ -462,7 +462,7 @@ async function main() {
nodes = rows.map(row => ({
id: row.id,
title: row.title,
notes: row.notes ?? null,
source: row.source ?? null,
description: row.description ?? null,
link: row.link ?? null,
dimensions: JSON.parse(row.dimensions_json || '[]'),
@@ -482,7 +482,7 @@ async function main() {
const words = trimmedQuery.split(/\s+/).filter(w => w.length > 0);
let sql = `
SELECT n.id, n.title, n.description, n.notes, n.link,
SELECT n.id, n.title, n.description, n.source, n.link,
n.created_at, n.updated_at, n.event_date,
COALESCE((SELECT JSON_GROUP_ARRAY(d.dimension)
FROM node_dimensions d WHERE d.node_id = n.id), '[]') as dimensions_json
@@ -492,7 +492,7 @@ async function main() {
const params = [];
for (const word of words) {
sql += ` AND (n.title LIKE ? COLLATE NOCASE OR n.description LIKE ? COLLATE NOCASE OR n.notes LIKE ? COLLATE NOCASE)`;
sql += ` AND (n.title LIKE ? COLLATE NOCASE OR n.description LIKE ? COLLATE NOCASE OR n.source LIKE ? COLLATE NOCASE)`;
params.push(`%${word}%`, `%${word}%`, `%${word}%`);
}
@@ -518,7 +518,7 @@ async function main() {
nodes = rows.map(row => ({
id: row.id,
title: row.title,
notes: row.notes ?? null,
source: row.source ?? null,
description: row.description ?? null,
link: row.link ?? null,
dimensions: JSON.parse(row.dimensions_json || '[]'),
@@ -560,13 +560,13 @@ async function main() {
for (const id of uniqueIds) {
const node = nodeService.getNodeById(id);
if (node) {
const rawChunk = node.chunk ?? null;
const rawChunk = node.source ?? node.chunk ?? null;
const chunkTruncated = rawChunk ? rawChunk.length > CHUNK_LIMIT : false;
nodes.push({
id: node.id,
title: node.title,
notes: node.notes ?? null,
source: node.source ?? node.notes ?? null,
description: node.description ?? null,
link: node.link ?? null,
chunk: chunkTruncated ? rawChunk.substring(0, CHUNK_LIMIT) : rawChunk,
@@ -610,14 +610,18 @@ async function main() {
throw new Error(descriptionError);
}
// Map MCP 'content' field → internal 'notes' field
// Map MCP legacy fields to canonical source
const mappedUpdates = { ...updates };
if (mappedUpdates.content !== undefined) {
mappedUpdates.notes = mappedUpdates.content;
delete mappedUpdates.content;
mappedUpdates.source = mappedUpdates.content;
}
if (mappedUpdates.chunk !== undefined && mappedUpdates.source === undefined) {
mappedUpdates.source = mappedUpdates.chunk;
}
delete mappedUpdates.content;
delete mappedUpdates.chunk;
const node = nodeService.updateNode(id, mappedUpdates, { appendNotes: true });
const node = nodeService.updateNode(id, mappedUpdates, { appendNotes: false });
return {
content: [{ type: 'text', text: `Updated node #${id}` }],
@@ -9,7 +9,7 @@ function getNodes(filters = {}) {
const { dimensions, search, limit = 100, offset = 0 } = filters;
let sql = `
SELECT n.id, n.title, n.description, n.notes, n.link, n.event_date, n.metadata, n.chunk,
SELECT n.id, n.title, n.description, n.source, n.notes, n.link, n.event_date, n.metadata, n.chunk,
n.created_at, n.updated_at,
COALESCE((SELECT JSON_GROUP_ARRAY(d.dimension)
FROM node_dimensions d WHERE d.node_id = n.id), '[]') as dimensions_json
@@ -30,7 +30,7 @@ function getNodes(filters = {}) {
// Text search
if (search) {
sql += ` AND (n.title LIKE ? COLLATE NOCASE OR n.description LIKE ? COLLATE NOCASE OR n.notes LIKE ? COLLATE NOCASE)`;
sql += ` AND (n.title LIKE ? COLLATE NOCASE OR n.description LIKE ? COLLATE NOCASE OR n.source LIKE ? COLLATE NOCASE)`;
params.push(`%${search}%`, `%${search}%`, `%${search}%`);
}
@@ -70,7 +70,7 @@ function getNodes(filters = {}) {
*/
function getNodeById(id) {
const sql = `
SELECT n.id, n.title, n.description, n.notes, n.link, n.event_date, n.metadata, n.chunk,
SELECT n.id, n.title, n.description, n.source, n.notes, n.link, n.event_date, n.metadata, n.chunk,
n.created_at, n.updated_at,
COALESCE((SELECT JSON_GROUP_ARRAY(d.dimension)
FROM node_dimensions d WHERE d.node_id = n.id), '[]') as dimensions_json
@@ -108,11 +108,10 @@ function createNode(nodeData) {
const {
title: rawTitle,
description,
notes,
source,
link,
event_date,
dimensions = [],
chunk,
metadata = {}
} = nodeData;
@@ -121,30 +120,23 @@ function createNode(nodeData) {
const now = new Date().toISOString();
const db = getDb();
// Build chunk fallback from available fields when no explicit chunk provided
let chunkToStore = chunk ?? null;
if (!chunkToStore || !chunkToStore.trim()) {
const fallbackParts = [title, description, notes].filter(Boolean);
const fallback = fallbackParts.join('\n\n').trim();
if (fallback) {
chunkToStore = fallback;
}
}
const sourceToStore = source && source.trim()
? source
: [title, description].filter(Boolean).join('\n\n').trim() || null;
const nodeId = transaction(() => {
const stmt = db.prepare(`
INSERT INTO nodes (title, description, notes, link, event_date, metadata, chunk, created_at, updated_at)
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
INSERT INTO nodes (title, description, source, link, event_date, metadata, created_at, updated_at)
VALUES (?, ?, ?, ?, ?, ?, ?, ?)
`);
const result = stmt.run(
title,
description ?? null,
notes ?? null,
sourceToStore,
link ?? null,
event_date ?? null,
JSON.stringify(metadata),
chunkToStore,
now,
now
);
@@ -169,11 +161,11 @@ function createNode(nodeData) {
/**
* Update an existing node.
* Note: notes is APPENDED by default (MCP tool behavior), not replaced.
* Source-first update path.
*/
function updateNode(id, updates, options = {}) {
const { appendNotes = true } = options;
const { title, description, notes, link, event_date, dimensions, chunk, metadata } = updates;
const { title, description, source, notes, link, event_date, dimensions, chunk, metadata } = updates;
const now = new Date().toISOString();
const db = getDb();
@@ -195,15 +187,15 @@ function updateNode(id, updates, options = {}) {
setFields.push('description = ?');
params.push(description);
}
if (notes !== undefined) {
if (appendNotes && existing.notes) {
// Append to existing notes
setFields.push('notes = ?');
params.push(existing.notes + '\n\n' + notes);
} else {
setFields.push('notes = ?');
params.push(notes);
}
if (source !== undefined) {
setFields.push('source = ?');
params.push(source);
} else if (notes !== undefined) {
const nextSource = appendNotes && existing.source
? `${existing.source}\n\n${notes}`
: notes;
setFields.push('source = ?');
params.push(nextSource);
}
if (link !== undefined) {
setFields.push('link = ?');
@@ -213,8 +205,8 @@ function updateNode(id, updates, options = {}) {
setFields.push('event_date = ?');
params.push(event_date);
}
if (chunk !== undefined) {
setFields.push('chunk = ?');
if (chunk !== undefined && source === undefined) {
setFields.push('source = ?');
params.push(chunk);
}
if (metadata !== undefined) {
+15 -11
View File
@@ -86,6 +86,7 @@ const sanitizeDimensions = (raw) => {
const addNodeInputSchema = {
title: z.string().min(1).max(160),
content: z.string().max(20000).optional(),
source: z.string().max(50000).optional(),
link: z.string().url().optional(),
description: z.string().max(2000).optional(),
dimensions: z.array(z.string()).min(1).max(5),
@@ -112,7 +113,7 @@ const searchNodesOutputSchema = {
z.object({
id: z.number(),
title: z.string(),
notes: z.string().nullable(),
source: z.string().nullable(),
description: z.string().nullable(),
link: z.string().nullable(),
dimensions: z.array(z.string()),
@@ -151,7 +152,7 @@ const getNodesOutputSchema = {
z.object({
id: z.number(),
title: z.string(),
notes: z.string().nullable(),
source: z.string().nullable(),
link: z.string().nullable(),
dimensions: z.array(z.string()),
updated_at: z.string()
@@ -349,7 +350,7 @@ mcpServer.registerTool(
inputSchema: addNodeInputSchema,
outputSchema: addNodeOutputSchema
},
async ({ title, content, link, description, dimensions, metadata, chunk }) => {
async ({ title, content, source, link, description, dimensions, metadata, chunk }) => {
const normalizedDimensions = sanitizeDimensions(dimensions);
if (normalizedDimensions.length === 0) {
throw new McpError(
@@ -360,12 +361,11 @@ mcpServer.registerTool(
const payload = {
title: title.trim(),
notes: content?.trim() || undefined,
source: source?.trim() || chunk?.trim() || content?.trim() || undefined,
link: link?.trim() || undefined,
description: description?.trim() || undefined,
dimensions: normalizedDimensions,
metadata: metadata || {},
chunk: chunk?.trim() || undefined
metadata: metadata || {}
};
const result = await callRaHApi('/api/nodes', {
@@ -422,7 +422,7 @@ mcpServer.registerTool(
nodes: nodes.map((node) => ({
id: node.id,
title: node.title,
notes: node.notes ?? null,
source: node.source ?? node.notes ?? null,
description: node.description ?? null,
link: node.link ?? null,
dimensions: node.dimensions || [],
@@ -446,12 +446,16 @@ mcpServer.registerTool(
throw new McpError(ErrorCode.InvalidParams, 'At least one field must be provided in updates.');
}
// Map MCP 'content' field → internal 'notes' field
// Map MCP legacy fields to canonical source
const mappedUpdates = { ...updates };
if (mappedUpdates.content !== undefined) {
mappedUpdates.notes = mappedUpdates.content;
delete mappedUpdates.content;
mappedUpdates.source = mappedUpdates.content;
}
if (mappedUpdates.chunk !== undefined && mappedUpdates.source === undefined) {
mappedUpdates.source = mappedUpdates.chunk;
}
delete mappedUpdates.content;
delete mappedUpdates.chunk;
const result = await callRaHApi(`/api/nodes/${id}`, {
method: 'PUT',
@@ -492,7 +496,7 @@ mcpServer.registerTool(
nodes.push({
id: result.node.id,
title: result.node.title,
notes: result.node.notes ?? null,
source: result.node.source ?? result.node.notes ?? null,
link: result.node.link ?? null,
dimensions: result.node.dimensions || [],
updated_at: result.node.updated_at
+13 -9
View File
@@ -35,6 +35,7 @@ const STATUS_PATH = path.join(
const addNodeInputSchema = {
title: z.string().min(1).max(160),
content: z.string().max(20000).optional(),
source: z.string().max(50000).optional(),
link: z.string().url().optional(),
description: z.string().max(2000).optional(),
dimensions: z.array(z.string()).min(1).max(5),
@@ -322,7 +323,7 @@ server.registerTool(
inputSchema: addNodeInputSchema,
outputSchema: addNodeOutputSchema
},
async ({ title, content, link, description, dimensions, metadata, chunk }) => {
async ({ title, content, source, link, description, dimensions, metadata, chunk }) => {
const normalizedDimensions = sanitizeDimensions(dimensions);
if (normalizedDimensions.length === 0) {
throw new Error('At least one dimension/tag is required when creating a node.');
@@ -330,12 +331,11 @@ server.registerTool(
const payload = {
title: title.trim(),
notes: content?.trim() || undefined,
source: source?.trim() || chunk?.trim() || content?.trim() || undefined,
link: link?.trim() || undefined,
description: description?.trim() || undefined,
dimensions: normalizedDimensions,
metadata: metadata || {},
chunk: chunk?.trim() || undefined
metadata: metadata || {}
};
const result = await callRaHApi('/api/nodes', {
@@ -393,7 +393,7 @@ server.registerTool(
nodes: nodes.map((node) => ({
id: node.id,
title: node.title,
notes: node.notes ?? null,
source: node.source ?? node.notes ?? null,
description: node.description ?? null,
link: node.link ?? null,
dimensions: node.dimensions || [],
@@ -417,12 +417,16 @@ server.registerTool(
throw new Error('At least one field must be provided in updates.');
}
// Map MCP 'content' field → internal 'notes' field
// Map MCP legacy fields to canonical source
const mappedUpdates = { ...updates };
if (mappedUpdates.content !== undefined) {
mappedUpdates.notes = mappedUpdates.content;
delete mappedUpdates.content;
mappedUpdates.source = mappedUpdates.content;
}
if (mappedUpdates.chunk !== undefined && mappedUpdates.source === undefined) {
mappedUpdates.source = mappedUpdates.chunk;
}
delete mappedUpdates.content;
delete mappedUpdates.chunk;
const result = await callRaHApi(`/api/nodes/${id}`, {
method: 'PUT',
@@ -463,7 +467,7 @@ server.registerTool(
nodes.push({
id: result.node.id,
title: result.node.title,
notes: result.node.notes ?? null,
source: result.node.source ?? result.node.notes ?? null,
link: result.node.link ?? null,
dimensions: result.node.dimensions || [],
updated_at: result.node.updated_at