profit/lakehouse
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: align search_workers + match_contract + /match to ethereal_workers_v1

Browse Source 
Same index↔dataset misalignment as the /search fix (423400e), in three more paths that
pair the ethereal_workers SQL dataset with the workers_500k_v1 vector index:
- search_workers MCP tool
- match_contract MCP tool
- /match HTTP handler

All now default to ethereal_workers_v1 (same 10K id-space as the SQL table). Verified:
/match role=Forklift Operator state=IL returns only IL forklift operators (sql_matches 94).
rag_question / /ask left as-is — pure RAG, no SQL join, so no id-space to misalign.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Claw 2026-06-30 22:44:44 -05:00
parent 423400e843
commit 35a80a7306
1 changed files with 7 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
mcp-server/index.ts
View File

@ -151,7 +151,9 @@ server.tool(
}, },
async ({ question, sql_filter, dataset, id_column, top_k }) => { async ({ question, sql_filter, dataset, id_column, top_k }) => {
const body: any = { const body: any = {
question, index_name: "workers_500k_v1", filter_dataset: dataset, id_column, top_k, generate: true, // Index must share an id-space with filter_dataset (default ethereal_workers, 10K)
// → ethereal_workers_v1. workers_500k_v1 (50K, different workers) misaligned.
question, index_name: "ethereal_workers_v1", filter_dataset: dataset, id_column, top_k, generate: true,
use_playbook_memory: true, use_playbook_memory: true,
}; };
if (sql_filter) body.sql_filter = sql_filter; if (sql_filter) body.sql_filter = sql_filter;
@ -185,7 +187,8 @@ server.tool(
if (city) filter += ` AND city = '${city}'`; if (city) filter += ` AND city = '${city}'`;
const r = await api("POST", "/vectors/hybrid", { const r = await api("POST", "/vectors/hybrid", {
question: `Find the best ${role} workers with relevant skills and certifications`, question: `Find the best ${role} workers with relevant skills and certifications`,
index_name: "workers_500k_v1", sql_filter: filter, // ethereal_workers_v1 aligns with filter_dataset ethereal_workers (same 10K ids).
index_name: "ethereal_workers_v1", sql_filter: filter,
filter_dataset: "ethereal_workers", id_column: "worker_id", filter_dataset: "ethereal_workers", id_column: "worker_id",
top_k: headcount * 2, generate: false, top_k: headcount * 2, generate: false,
use_playbook_memory: true, use_playbook_memory: true,
@ -556,7 +559,8 @@ async function main() {
if (b.city) filter += ` AND city = '${b.city}'`; if (b.city) filter += ` AND city = '${b.city}'`;
return ok(await api("POST", "/vectors/hybrid", { return ok(await api("POST", "/vectors/hybrid", {
question: `Best ${b.role} workers with relevant skills`, question: `Best ${b.role} workers with relevant skills`,
index_name: b.index || "workers_500k_v1", sql_filter: filter, // Align default index with the ethereal_workers dataset (see /search fix).
index_name: b.index || "ethereal_workers_v1", sql_filter: filter,
filter_dataset: b.dataset || "ethereal_workers", filter_dataset: b.dataset || "ethereal_workers",
id_column: "worker_id", top_k: (b.headcount || 5) * 2, generate: false, id_column: "worker_id", top_k: (b.headcount || 5) * 2, generate: false,
use_playbook_memory: true, use_playbook_memory: true,