Content Systems
Content Systems
Five content pillars capture, codify, and retrieve knowledge across the domus-captures component. Each pillar has its own naming convention, directory location, governing standard, and retrieval pattern.
Worklogs
The daily capture system.
Every working day produces one worklog via make new-day, which scaffolds the file and section partials automatically.
Naming: WRKLOG-YYYY-MM-DD.adoc
Section partials (8 files in partials/worklog/):
| Partial | Purpose |
|---|---|
|
Time-sensitive items requiring immediate action |
|
Morning planning and priority setting |
|
Workplace tasks and progress |
|
Personal infrastructure and project work |
|
Study sessions, reading, certification progress |
|
Home lab and infrastructure changes |
|
CLI commands worth preserving (feeds the Codex) |
|
Cross-references to patterns, standards, and case studies |
Volume: 105 worklogs in 2026 as of early April — roughly one per working day.
Workflow:
make new-day # Scaffold today's WRKLOG with all section includes
# Write content into the section partials
make serve # Build and preview at localhost:8000
Pattern Journal
The field notebook. Patterns capture practitioner knowledge using an Aristotelian taxonomy that distinguishes types of knowing.
Taxonomy:
| Category | Greek | What It Captures |
|---|---|---|
Techne |
Craft knowledge |
How to do things — procedures, techniques, tool usage |
Episteme |
Theoretical knowledge |
Why things work — principles, models, architectures |
Phronesis |
Practical wisdom |
When to apply which approach — judgment, trade-offs |
Poiesis |
Creative production |
Making new things — scripts, configurations, documentation |
Scale: 95 pages across 18 domains including networking, security, CLI, documentation, infrastructure, and workflow.
Format: Dated entries with context headers. Each entry records the situation, the insight, and the generalized principle. Patterns are descriptive ("I noticed…"), never prescriptive — that role belongs to standards.
Philosophy: The pattern journal functions as a field notebook in the naturalist tradition. Observations accumulate. When a pattern recurs across three or more entries, it becomes a candidate for promotion to a standard.
Codex ("The Arsenal")
The CLI reference library. The Codex is the quick-lookup system for commands, syntax, and tool-specific techniques.
Scale: 110 pages across 15 categories.
Structure: Code lives in examples/ and is included via include::example$codex/filename.ext[].
Partials in partials/codex/ hold the explanatory text.
Pages in pages/codex/ are thin shells.
Governance: STD-016 defines the Codex format:
-
Title states the tool and operation
-
Synopsis shows the minimal invocation
-
Options table covers flags used in practice (not exhaustive man-page replication)
-
Examples are executable — copy-paste must work
-
Related section links to patterns where the command appears in context
Categories include: awk, sed, grep, find, xargs, git, ssh, systemd, jq, curl, openssl, vault, nmcli, firewall-cmd, and package management.
Case Studies
The incident and change record system. Case studies document real events with enough detail to reconstruct the timeline, understand the root cause, and prevent recurrence.
Scale: 76 files across 5 types.
Taxonomy (governed by STD-013):
| Prefix | Type | Purpose |
|---|---|---|
|
Vendor TAC Case |
Escalations to vendor support with case numbers and resolution |
|
Incident Report |
Unplanned outages or security events with timeline |
|
Change Request |
Planned changes with phases, rollback, and validation |
|
Root Cause Analysis |
Post-incident deep dive with prevention measures |
|
Deployment Record |
New system deployments with validation checklists |
Cross-reference chain: Case studies do not exist in isolation. The expected linkage is:
INC (what happened) ──▶ RCA (why) ──▶ Pattern (generalized) ──▶ Standard (codified)
Every incident SHOULD produce an RCA. Every RCA SHOULD identify or update at least one pattern. Patterns that recur across three RCAs SHOULD be promoted to standards.
Sessions
Claude Code collaboration captures. Sessions document AI-assisted work with enough detail to reproduce the approach and teach the techniques.
Naming: SESSION-YYYY-MM-DD-<slug>.adoc
Content:
| Section | Purpose |
|---|---|
Deliverables |
What was produced — files created, configs changed, commits made |
Commits |
Git log of all commits made during the session |
Agents & Modes |
Which Claude Code agents and modes were used and why |
Teaching Points |
CLI techniques, architectural decisions, and patterns learned |
Decisions |
Trade-offs considered and choices made with rationale |
Sessions serve dual purposes: they are a work record (what happened) and a curriculum artifact (what was taught). The teaching-points section is the most valuable for long-term retrieval.