Nest From-Scratch Ontology
This note captures a from-scratch reframing of what Vulture Nest is trying to be at its most basic level.
The main shift is:
Vulture Nest should be understood less as a wiki that contains notes and more as a persistent cognitive environment for humans and agents.
That means notes are not the center of the system. They are one artifact class among several.
Why This Reframe Matters
The current Nest already does more than store knowledge.
It also:
- coordinates work across sessions
- preserves seams and handoffs
- supports dialogue and iterative reasoning
- generates outputs through tools and execution
- turns interaction into durable state
That combination suggests the Nest is not just:
- a repository
- a workflow tracker
- a wiki
It is a continuity system for ongoing cognition and execution.
The Core Modes
At minimum, the Nest appears to operate in three modes:
1. Artifact Mode
The system creates, edits, links, and preserves artifacts.
Examples:
- permanent notes
- literature notes
- fleeting notes
- handoffs
- logs
- generated files
2. Interaction Mode
The system hosts live inquiry through discussion, debate, prompting, and back-and-forth refinement.
Examples:
- human-agent conversations
- turn-based debate artifacts
- review loops
- synthesis passes
3. Execution Mode
The system runs actions that test, transform, or produce state.
Examples:
- scripts
- builds
- validation passes
- indexing
- REPL-like experimentation
The Deeper Primitives
If the Nest were designed from scratch around what it actually does, the primitive entities would likely be:
- artifact
- interaction
- execution
- memory
- state
- transition
- provenance
- continuation
These are more fundamental than "note" alone.
What The System Really Has To Answer
A from-scratch Nest would need to answer:
- what enters the system?
- what kinds of things can exist in it?
- what transformations are allowed?
- what becomes durable?
- what remains provisional?
- what expires?
- what can be resumed later?
- what is authored directly?
- what is computed or derived?
Those questions are more basic than note taxonomy by itself.
Proposed Core Ontology
Artifact
A durable or semi-durable unit of recorded state.
Examples:
- note
- handoff
- seam
- log entry
- generated report
Interaction
A live exchange that produces interpretation, challenge, or decision.
Examples:
- conversation
- debate turn
- review
- synthesis pass
Execution
An action taken against the environment that produces evidence or changes state.
Examples:
- script run
- validator pass
- crawl
- build
- REPL experiment
Memory
The retained portion of system state that is meant to persist and be reused.
Examples:
- durable wiki notes
- indexed source records
- session traces worth preserving
Transition
A rule or pathway by which one thing becomes another.
Examples:
- source intake -> literature note
- literature note -> permanent note
- active discussion -> RFC
- session seam -> resumed work
- experiment -> accepted result or discard
Continuation
The ability for work, thought, or context to be resumed by a human or another agent without starting over.
This may be the Nest's deepest function.
The Key Design Insight
The important thing is not just that artifacts exist.
The important thing is that:
- interactions generate artifacts
- artifacts constrain later interactions
- executions test or change the shared state
- memory preserves what should survive
- continuation lets work move across time, sessions, and agents
That loop is closer to the Nest's actual essence than "a wiki with notes."
If Starting Over Entirely
The cleaner aim would be:
Build a persistent cognitive environment for humans and agents, where artifacts, interactions, and executions share one continuity layer.
From that aim, notes become one part of a larger system instead of the foundational assumption.
Implications For Future Spec Work
This reframing suggests future specs should distinguish at least:
- artifact schema
- interaction protocol
- execution surfaces
- memory/promotion rules
- continuation and handoff rules
It also suggests that frontmatter and note taxonomy alone will never fully model the Nest, because some of its core behavior lives in interaction and execution, not only in documents.