Why start here?
The moment a system should learn from real use, the same risk appears: optimization becomes hidden, unverifiable, or too eager. Dobby is the layer that allows learning without bypassing governance and approval.
Adaptation
The part that makes tomorrow better than today
Controlled adaptation and safe system learning.
Dobby improves the system from real evidence, but only in a controlled, explainable, approval-aware way.
Improvement without losing control
Status Available now README sync 22 May 2026
When do I need this?
Dobby becomes important when improvement is not only desired, but must remain explainable and bounded in live operations.
- when runtime traces need to become replayable learning signals
- when adaptation is only acceptable with approval, replay, and evidence
- when optimization must fail closed instead of silently writing itself into the stack
What role it plays here
Dobby is not the system that mutates autonomously. It is the adaptive-learning layer that forces evidence, replay, proposals, and approval into one controlled path.
What the module actually does
Its value is not learning by itself, but making sure learning remains explainable, replayable, and stoppable afterwards.
At the core
Accepts learning signals in a governed way. Dobby does not collect runtime signals as hunches, but as structured, tenant-aware, guardrail-bound intake.
Evaluates candidates through replay instead of hope. Potential improvements are not guessed live; they are checked against replay, threshold, and contract logic.
Keeps approval and proposal lifecycle together. Dobby separates signal, evaluation, proposal, and activation so improvement cannot slip past governance in secret.
Degrades fail-closed. When Fabric truth, approval, or persistence are missing, learning does not turn into improvisation; it turns into a controlled stop.
What role it plays in the stack
Within the stack, Dobby is where improvement becomes accountable. Other modules provide truth, approvals, evidence, or provenance; Dobby makes sure that does not turn into uncontrolled self-modification.
uses Fabric as governance and contract truth instead of inventing local rules
reads approval truth from Warp instead of legitimizing activation itself
connects learning evidence to Shuttle/Bobbin signals without taking over their ownership
What this looks like in practice
Dobby proves its value when improvement stays explainable instead of disappearing as silent mutation inside the system.
Learning traces arrive cleanly
Runtime signals are accepted as governed intake instead of being collected as loose heuristics.
Replay separates a good idea from a viable candidate
Improvements are checked against contracts, thresholds, and repeatability before they move forward.
Proposal and approval stay visibly separate
A candidate is never mistaken for an approval; every activation stays bound to approval truth.
Under drift, learning stops fail-closed
Missing truth or broken persistence does not trigger hidden optimization; it triggers controlled degradation.
How it fits into the system
Dobby does not stand alone. It connects to neighboring modules so a single capability becomes dependable follow-through.
Important boundary
Dobby stays bounded to its role as Controlled adaptation and safe system learning. It does not replace other modules; it makes its part of the system traceable, connectable, and reviewable.
What comes next
Bobbin Persistent memory and provenance layer. Tenter Voice runtime, Asterisk operations, and provable operator verification. Pattern Consistency and control layer against drift. Warp Runtime orchestration and controlled execution steering. Beam Safety, upgrade, and change-verification layer. All modules Back to overview
What is intentionally out of scope
Dobby is intentionally learning and proposal logic, not a second autonomous control plane.
no separate contract or admission truth next to Fabric
no autonomous activation without an approval lane
no hidden repo or runtime writeback outside governed proposal paths
What keeps this page honest
This explanation stays anchored to the module’s current truth, including its real boundaries, responsibilities, and contracts.
Dobby is the adaptation layer where the system gets better from real evidence without losing control.
README.md
Source and repo truth
This page is rendered from the repo-owned projection truth and remains tied to the README, module boundaries, and status.
GitHub JaddaHelpifyr/jhf-dobbyDobby
Within the stack, Dobby is where improvement becomes accountable. Other modules provide truth, approvals, evidence, or provenance; Dobby makes sure that does not turn into uncontrolled self-modification.