Skip to main content

​
When to Use

At the end of any work session that produced learnings, technical findings, or campaign content worth preserving. The archive system distributes content across 5 repos based on content type.

​
The 5-Repo Network

RepoAcceptsPath TemplateAudience
devarno-cloud/atlaslearning, campaignlearnings/{date}-{slug}.md, campaigns/{date}-{slug}.mdbusiness, marketing
so1-io/so1-contentfindingfindings/{date}-{slug}.mdengineering
so1-io/veritaspromptquarantine pipelineengineering
traceo-ai/arielbaseline, specbaseline/, specs/engineering
traceo-ai/traceo-mcp-server— (read-only)receives via ariel pip dep—

​
File Naming

{YYYY-MM-DD}-{slug}.md — e.g., 2026-03-28-schema-first-foundation-before-cli.md

​
Required Frontmatter

---
title: "Descriptive title"
date: 2026-03-28
type: learning  # or: finding, campaign, prompt, baseline, spec
tags:
  - relevant-tags
source_session: archive-2026-03-28-{description}
cross_ref:
  - repo: so1-io/so1-content
    path: findings/2026-03-28-corresponding-finding.md
    type: finding
---
All documents produced in the same archive invocation must cross-reference each other via the cross_ref array.

​
Content Type Definitions

  • Learning: Business-facing takeaway. Max 500 words. “What did we learn that affects product, strategy, or GTM?”
  • Finding: Technical discovery. Max 500 words. “What bug, pattern, or decision emerged from engineering?”
  • Campaign: Marketing content seed. Max 500 words. “What messaging angles can we extract for design journal, Show HN, LinkedIn?”

​
Workflow

  1. Identify archivable content from the session (learnings, findings, campaigns)
  2. Write each document with full frontmatter and cross-references
  3. Stage and commit to each repo separately
  4. Do NOT push — commits are local only

​
Local Repo Paths

/home/devarno/code/workspace/atlas/                  (devarno-cloud/atlas)
/home/devarno/code/workspace/so1-io/so1-content/
/home/devarno/code/workspace/so1-io/veritas/
/home/devarno/code/workspace/ariel/