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

fix: align 4 more intelligence routes to ethereal_workers_v1

Browse Source 
Same index↔dataset misalignment swept across the intelligence handler — four more
/vectors/hybrid calls that paired workers_500k_v1 with filter_dataset ethereal_workers:
- similar-workers route (~2418)
- semantic role-discovery route (~2459)
- smart_search route (~2869)
- scenario worker-find (~3531)

Verified: no workers_500k_v1 index remains paired with the ethereal_workers dataset;
the 9 remaining refs are correctly aligned (workers_500k dataset) or pure RAG. Server
boots clean. Takes effect on next mcp-server restart.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Claw 2026-06-30 22:47:16 -05:00
parent 35a80a7306
commit 15703063ea
1 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
mcp-server/index.ts
View File

@ -2415,7 +2415,7 @@ async function main() {
queries.push(`Vector: Semantic similarity on ${worker.name}'s full profile → ${stateFilter}`); queries.push(`Vector: Semantic similarity on ${worker.name}'s full profile → ${stateFilter}`);
const searchR = await api("POST", "/vectors/hybrid", { const searchR = await api("POST", "/vectors/hybrid", {
question: worker.resume_text || `${worker.role} in ${worker.city} with skills ${worker.skills}`, question: worker.resume_text || `${worker.role} in ${worker.city} with skills ${worker.skills}`,
index_name: "workers_500k_v1", index_name: "ethereal_workers_v1", // align with ethereal_workers dataset
sql_filter: stateFilter + ` AND CAST(reliability AS DOUBLE) >= 0.7`, sql_filter: stateFilter + ` AND CAST(reliability AS DOUBLE) >= 0.7`,
filter_dataset: "ethereal_workers", id_column: "worker_id", top_k: 5, generate: false, filter_dataset: "ethereal_workers", id_column: "worker_id", top_k: 5, generate: false,
}); });
@ -2456,7 +2456,7 @@ async function main() {
queries.push(`Vector: Semantic search for "${roleDesc}" — no exact role match needed`); queries.push(`Vector: Semantic search for "${roleDesc}" — no exact role match needed`);
const searchR = await api("POST", "/vectors/hybrid", { const searchR = await api("POST", "/vectors/hybrid", {
question: `Worker experienced in ${roleDesc}, relevant skills and certifications`, question: `Worker experienced in ${roleDesc}, relevant skills and certifications`,
index_name: "workers_500k_v1", sql_filter: "CAST(reliability AS DOUBLE) >= 0.75", index_name: "ethereal_workers_v1", sql_filter: "CAST(reliability AS DOUBLE) >= 0.75", // align w/ dataset
filter_dataset: "ethereal_workers", id_column: "worker_id", top_k: 8, generate: false, filter_dataset: "ethereal_workers", id_column: "worker_id", top_k: 8, generate: false,
}); });
return ok({ type: "discovery", summary: `${(searchR.sources||[]).length} workers found through semantic skill matching for: "${roleDesc}"`, return ok({ type: "discovery", summary: `${(searchR.sources||[]).length} workers found through semantic skill matching for: "${roleDesc}"`,
@ -2866,7 +2866,7 @@ async function main() {
const [searchR, directR, patternR] = await Promise.all([ const [searchR, directR, patternR] = await Promise.all([
api("POST", "/vectors/hybrid", { api("POST", "/vectors/hybrid", {
question: q, index_name: "workers_500k_v1", sql_filter: filterStr, question: q, index_name: "ethereal_workers_v1", sql_filter: filterStr, // align w/ dataset
filter_dataset: "ethereal_workers", id_column: "worker_id", filter_dataset: "ethereal_workers", id_column: "worker_id",
// Honor the parsed headcount (capped at 25 to keep the // Honor the parsed headcount (capped at 25 to keep the
// vector rerank from re-scoring more rows than render). // vector rerank from re-scoring more rows than render).
@ -3528,7 +3528,7 @@ async function runWeekSimulation() {
const filt = `role = '${role}' AND state = '${state}' AND reliability >= ${minRel}`; const filt = `role = '${role}' AND state = '${state}' AND reliability >= ${minRel}`;
const r = await api("POST", "/vectors/hybrid", { const r = await api("POST", "/vectors/hybrid", {
question: `Find ${role} workers in ${city}, ${state} for ${scenario.situation}`, question: `Find ${role} workers in ${city}, ${state} for ${scenario.situation}`,
index_name: "workers_500k_v1", index_name: "ethereal_workers_v1", // align with ethereal_workers dataset
sql_filter: filt, sql_filter: filt,
filter_dataset: "ethereal_workers", filter_dataset: "ethereal_workers",
id_column: "worker_id", id_column: "worker_id",