Research Log Entry — Corpus Assembly Method

---

title: "Corpus Assembly Method" author: Mark Rabideau, Khoj, Ardens Braid (Human + AI) date: September 2025

---

Context

The Ardens Project requires a unified text corpus (ardens_corpus.txt) to serve as a shared substrate for retrieval, synthesis, and reflexive analysis. Since project materials are authored across multiple Markdown files, a reproducible process is needed to assemble these fragments into a single, analyzable document.

---

Method

A Bash script is used to concatenate all Markdown files in the repository into a single corpus file. To preserve provenance and maintain deterministic ordering, each file is prefixed with a marker line and processed in sorted order.

---

Script

```bash

!/bin/bash

Corpus Assembly Script

Concatenates all Markdown files into a unified corpus

Adds file markers for traceability and uses sorted order for reproducibility

output="ardens_corpus.txt" rm -f "$output"

find . -type f -name "*.md" | sort | while read file; do echo "===== $file =====" >> "$output" cat "$file" >> "$output" echo -e "\n" >> "$output" done