heimdall

A lightweight CPU only memory approach with ranked retrieval. Simple, yet effective

A new way for AI to search for data
View on GitHub
136GitHub stars

Python, JavaScript

What XERJ indexed

Snapshot

Whole repository

The whole repository at commit 7e38b721a7 (2026-10-23).

Selected work

Deva
@ArihantDeva

“cut kb_search 21.8s -> 159ms with a warm fused worker”. 8 files, 1,449 lines in bin, tests.

Licence

The repository

MIT

Checked by a person at the pinned commit.

The XERJ corpus

MIT, with credit

The author agreed by email on 2026-10-23. Same terms as the repository. Answers name the repo, the commit and the file:line.

in licence reviewnot in the hub yetsuggested use: adapt with attribution

ai-agentsclaude-codeclicode-intelligencedeveloper-toolsembeddingsknowledge-basellm-tools

Example questions

The prompt

Paste it into your AI agent. It clones ArihantDeva/heimdall at the pinned commit and indexes it on your machine.

Search the GitHub repository ArihantDeva/heimdall with XERJ (https://hub.xerj.org/corpus/arihantdeva--heimdall).
Run these in a shell, in order, and check each result before the next:
1. xerj --help  (success: it prints usage). If xerj is missing: curl -fsSL https://xerj.org/get | sh
2. git clone https://github.com/ArihantDeva/heimdall heimdall && git -C heimdall checkout 7e38b721a7fe696b4035c3d9de9f13ba75492728
3. Start a local node and leave it running: xerj --insecure --data-dir ./xerj-data
4. xerj autoindex ./heimdall --url http://127.0.0.1:9200 --prefix heimdall --progress plain
5. xerj autoindex map --url http://127.0.0.1:9200
Then answer my question from the indexed files and cite the ax_path of every hit. Query syntax: https://xerj.org/llms.txt
Licence: MIT. Keep the copyright notice and credit the authors.

Related

Selected work and consent: the XERJ registry.