Local semantic code search that saves your agent tokens

developer-tools / Show HN

49

SCORE

Reference is a local semantic code search tool designed for AI coding agents. It allows agents to search codebases by meaning rather than exact keywords, reducing the number of tokens consumed during context retrieval. The tool runs locally, keeping code private without sending it to external services.

VISIT PRODUCT →

Sources (1)

Show HN

https://news.ycombinator.com/item?id=49164316

1 PTS

Score Breakdown

Traction

raw 1.00 · weight 35%

3.5pts

Novelty

0 days old · weight 20%

20.0pts

Source diversity

1 source · weight 10%

3.3pts

AI quality

raw 62.00 · weight 35%

21.7pts

Clear value prop (local semantic search for AI agents, token savings, privacy), but zero traction signals and unproven execution limit confidence.

Final Score49/100

Similar Products

AI-TOOLS · GITHUB

DEVELOPER-TOOLS · GITHUB

DEVELOPER-TOOLS · GITHUB

DEVELOPER-TOOLS · GITHUB

DEVELOPER-TOOLS · GITHUB

DEVELOPER-TOOLS · GITHUB

← ALL PRODUCTS© 2026 MARKETHUNT