Skip to content

glossary.md

Latest commit

 

History

History
95 lines (61 loc) · 2.36 KB

glossary.md

File metadata and controls

95 lines (61 loc) · 2.36 KB

Loop Glossary

A

Activation Decision The classifier outcome that decides whether Loop should stay simple or invoke multi-step recursive orchestration.

B

Behavior (Topos) A semantic contract describing an operation, including preconditions, postconditions, and error cases.

Beads (bd) The issue tracking system used by this repository for planning, status, and closure.

C

Completeness Mode Spec-agent generation policy:

  • Baseline: placeholder-free stubs
  • Placeholder: explicit draft: annotations with executable stubs (no TODO/sorry tokens)

Context Externalization The process of transforming context into structured variables used by runtime orchestration and prompts.

D

Dual-Track Sync The synchronization layer between Topos semantic specs and Lean formal artifacts.

E

Epistemic Verification Claim-level confidence and consistency checks to reduce hallucination risk.

F

Formalization Level Depth of generated formal artifacts:

  • Types
  • Invariants
  • Contracts
  • FullProofs

G

Governance Gates (dp) Repository policy checks (review, verify, enforce pre-commit, enforce pre-push) executed via ./scripts/dp.

L

Lean REPL Interactive Lean execution surface used for checking generated or authored proof artifacts.

Loop Mode Execution path where recursive orchestration is activated for complex tasks.

M

Memory Store SQLite-backed hypergraph used for fact, experience, and reasoning trace persistence.

O

OODA Observe/Orient/Decide/Act loop used to reason about runtime behavior and validation flows.

P

Pattern Classifier The complexity detector used to determine whether orchestration should activate.

R

RLM (Recursive Language Model) The orchestration approach that decomposes large tasks into tractable sub-queries with shared context and synthesis.

Reasoning Trace Decision graph capturing options, chosen paths, actions, and outcomes.

S

Spec Agent Workflow engine that transforms natural language requirements into Topos + Lean artifacts.

SUBMIT Typed-signature output completion mechanism used to finalize structured module outputs.

T

Topos Human-readable semantic specification format (.tps) used for concepts, behaviors, and requirements.

Trajectory Events Structured event stream describing runtime execution progress and decisions.