# Cognitive OS Starter Filesystem Template

Use this as a minimal, plain-text operating structure for AI-native work. Keep the files short enough for agents to inspect and explicit enough for humans to audit.

```text
/cognitive-os
  /memory
    profile.md
    preferences.md
    decisions.md
    lessons.md
  /workflows
    research-loop.md
    writing-loop.md
    review-loop.md
  /agents
    researcher.md
    editor.md
    operator.md
  /projects
    project-name/
      brief.md
      sources.md
      drafts.md
      decisions.md
  /reviews
    weekly-review.md
    monthly-system-review.md
```

## `/memory/profile.md`

```md
# Profile

## Role / context
-

## Current goals
-

## Durable constraints
-

## What future agents should know
-
```

## `/memory/preferences.md`

```md
# Preferences

## Output style
-

## Decision preferences
-

## Things to avoid
-
```

## `/memory/decisions.md`

```md
# Decisions

## YYYY-MM-DD — Decision title

- Decision:
- Context:
- Alternatives considered:
- Why this decision stands:
- What would change it:
```

## `/memory/lessons.md`

```md
# Lessons

## YYYY-MM-DD — Lesson title

- What happened:
- What worked:
- What failed:
- What should change next time:
- File/workflow updated:
```

## `/workflows/research-loop.md`

```md
# Research Loop

## Purpose
-

## Inputs
-

## Steps
1.
2.
3.

## Output standard
-

## Human approval required for
-

## Update after each run
-
```

## `/workflows/writing-loop.md`

```md
# Writing Loop

## Purpose
-

## Inputs
-

## Drafting steps
1.
2.
3.

## Editorial standards
-

## Pre-publish checks
-

## Update after each run
-
```

## `/workflows/review-loop.md`

```md
# Review Loop

## Questions
- What did the agent produce?
- What mattered?
- What context was missing?
- What should become a file?
- What should be constrained?
- What is the next experiment?

## Required ending
Update at least one file, checklist, workflow, source map, decision log, or constraint.
```

## `/agents/researcher.md`

```md
# Researcher Agent

## Role
Find, compare, and evaluate sources.

## Must read before acting
-

## Allowed actions
-

## Not allowed without approval
-

## Output standard
-
```

## `/agents/editor.md`

```md
# Editor Agent

## Role
Improve structure, clarity, argument, and reader value.

## Must read before acting
-

## Allowed actions
-

## Not allowed without approval
-

## Output standard
-
```

## `/agents/operator.md`

```md
# Operator Agent

## Role
Execute operational tasks safely and verify outcomes.

## Must read before acting
-

## Allowed actions
-

## Not allowed without approval
-

## Verification standard
-
```

## `/projects/project-name/brief.md`

```md
# Project Brief

## Goal
-

## Why this matters
-

## Current state
-

## Constraints
-

## Success criteria
-

## Next action
-
```

## `/projects/project-name/sources.md`

```md
# Sources

## Source name

- URL/reference:
- Why it matters:
- Trust level:
- Notes:
```

## `/projects/project-name/decisions.md`

```md
# Project Decisions

## YYYY-MM-DD — Decision title

- Decision:
- Context:
- Tradeoffs:
- Revisit when:
```