Vaner

Policy Bundles

The seven shipped bundles, who each is for, and what it implies for compute, cloud, spend, and Deep-Run.

Ask ChatGPTAsk Claude

A policy bundle is the intermediate layer between user-facing setup and the engine's actual knobs. Each bundle names an outcome-level archetype (e.g. "Local Balanced", "Hybrid Quality") and carries the dimensions the engine consults: cloud posture, runtime profile, spend profile, latency profile, privacy profile, prediction-horizon bias, drafting aggressiveness, exploration ratio, persistence strength, goal weighting, context-injection default, and Deep-Run profile.

Bundles compose multiplicatively with Deep-Run preset overrides at session start. Bundles set the baseline; Deep-Run sessions layer additive overrides on top.

The single source of truth is vaner.setup.catalog.PROFILE_CATALOG. The seven shipped bundles are listed below in catalog order.

Local Lightweight

id: local_lightweight

Run entirely on this device with a small local model. Best for older laptops, battery, and privacy-first users.

Who it's for. Laptop on battery, integrated GPU, no cloud budget. You want Vaner not to heat the room.

DimensionValue
Local / cloud posturelocal_only
Runtime profilesmall
Spend profilezero
Latency profilesnappy
Privacy profilestrict
Drafting aggressiveness0.8 (cautious)
Exploration ratio-0.05
Persistence strength0.7 (decay fast)
Context injection defaultdigest_only
Deep-Run profileconservative

Local Balanced

id: local_balanced

Local-first with a mid-sized model. Solid default for a capable laptop or desktop with no cloud reliance.

Who it's for. Capable laptop or desktop, willing to ponder more in the background. The default for the Capable hardware tier.

DimensionValue
Local / cloud posturelocal_preferred
Runtime profilemedium
Spend profilezero
Latency profilebalanced
Privacy profilestrict
Drafting aggressiveness1.0
Exploration ratio0.0
Persistence strength1.0
Context injection defaultpolicy_hybrid
Deep-Run profilebalanced

Local Heavy

id: local_heavy

All-local with a large model and aggressive background pondering. Built for workstations and homelab GPUs.

Who it's for. Workstation or homelab with a large local model (e.g. 30–70B at Q4+). Local-only by policy, aggressive by capacity.

DimensionValue
Local / cloud posturelocal_only
Runtime profilelarge
Spend profilezero
Latency profilequality
Privacy profilestrict
Drafting aggressiveness1.1
Exploration ratio+0.10
Persistence strength1.3
Context injection defaultpolicy_hybrid
Deep-Run profileaggressive

Hybrid Balanced

id: hybrid_balanced

Local for everyday work, cloud only when it's clearly worth it. Recommended default for most users.

Who it's for. Anyone fine with the engine reaching for a cloud model when local is genuinely insufficient. The canonical default when the hardware tier is unknown, safe middle path.

DimensionValue
Local / cloud posturehybrid
Runtime profilemedium
Spend profilelow
Latency profilebalanced
Privacy profilestandard
Drafting aggressiveness1.0
Exploration ratio+0.05
Persistence strength1.0
Context injection defaultpolicy_hybrid
Deep-Run profilebalanced

Hybrid Quality

id: hybrid_quality

Reach for the best available model, local or cloud, and trade some snappiness and budget for higher answer quality.

Who it's for. Serious work where you'd rather pay than wait. Cloud bias is set higher; latency optimises for quality over snap.

DimensionValue
Local / cloud posturecloud_preferred
Runtime profilelarge
Spend profilemedium
Latency profilequality
Privacy profilestandard
Drafting aggressiveness1.2
Exploration ratio+0.10
Persistence strength1.2
Context injection defaulttop_match_auto_include
Deep-Run profilebalanced

Cost Saver

id: cost_saver

Hybrid setup with tight cloud spend caps. Prefers the cheapest endpoint that still meets the bar.

Who it's for. Cost-conscious user who still wants some hybrid capacity. Tight spend cap; cheapest endpoint that meets the latency and context floor.

DimensionValue
Local / cloud posturelocal_preferred
Runtime profilesmall
Spend profilelow
Latency profilebalanced
Privacy profilestandard
Drafting aggressiveness0.9
Exploration ratio-0.05
Persistence strength0.9
Context injection defaultadopted_package_only
Deep-Run profileconservative

Deep Research

id: deep_research

Long-horizon, broad exploration for research and writing. Persists predictions longer and prefers depth over speed.

Who it's for. Research / writing work where Vaner should ponder broadly and persistently. Overnight Deep-Run is the natural mode; daytime use composes the bundle's bias with the active session's preset.

DimensionValue
Local / cloud posturehybrid
Runtime profilelarge
Spend profilemedium
Latency profilequality
Privacy profilestandard
Drafting aggressiveness1.1
Exploration ratio+0.15 (most explore-leaning)
Persistence strength1.5 (slowest decay)
Goal weighting1.4 (highest)
Context injection defaultpolicy_hybrid
Deep-Run profileaggressive

How a bundle is selected

vaner setup recommend scores all seven bundles against detected hardware tier and returns the best match. The reasoning trail ("hybrid because cloud_posture = hybrid; medium runtime because tier = capable; balanced Deep-Run because preparation = standard") is surfaced verbatim in the cockpit.

To inspect or change the active bundle:

vaner setup show       # current bundle + reasons + hardware
vaner setup recommend  # JSON: candidate bundles with scores
vaner setup apply      # persist + hot-reload

On this page