vincitamore/amore-build/somniator

Claude Code and agent skill by vincitamore/amore-build

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Dream review and agentic-dream contracts for Lucerna. Use when reviewing dream manifests (review-status pending), applying or closing forge proposals, reading forge/dreams sessions, or running supervised self-orient / agentic-housekeeping cycles. Load before multi-turn dream review work. SKIP plain org CRUD; SKIP operator forge pipelines under forge/sessions/ (those are forge-master).

Install

npx vskill@latest install vincitamore/amore-build/somniator

SKILL.md

somniator - dream review for Lucerna

Lucerna may run autonomous light and agentic dreams when the house has dreamsEnabled set. This skill teaches a resident agent how to review those artifacts without treating them as auto-applied changes.

Path map

forge/
├── dreams/
│ ├── {ts}-{action}.md # light + agentic reports
│ └── sessions/
│ └── {YYYYMMDD-HHmmss}-{action}.manifest.md
└── proposals/
└── {slug}.md # status: pending | applied | closed
ArtifactPending signalResident verb
Light / agentic reportstatus: pendingRead findings; set status: acted when processed
Dream pipeline manifestreview-status: pendingReview body; set review-status: reviewed
Proposalstatus: pendingApply carefully or close; flip status

Manifest review

Manifest path shape:

forge/dreams/sessions/<YYYYMMDD-HHmmss>-<action>.manifest.md

Frontmatter includes type: forge, pipeline: dream-<action>, recipe: dream, triggered-by: dream, and review-status: pending.

What `review-status: pending` means: the cycle finished and left work for a resident. The body lists what ran, what was read, and what was produced. If a governance breach section is present, treat that as blocking: inspect the named paths, remediate by hand if needed, then review. Lucerna does not auto-revert out-of-bounds writes.

How to mark reviewed: edit the manifest frontmatter in place and set review-status: reviewed. Do not delete the manifest; keep the audit trail.

Proposal handling

Proposal path: forge/proposals/<slug>.md.

Frontmatter:

yaml
type: proposal
status: pending
created: 'YYYY-MM-DD'
triggered-by: dream
title: "<one line>"
target: <relative path>

Proposals are doc-only. Lucerna never applies them. A resident:

  1. Reads the proposed change and rationale.
  2. If accepting, applies the change under operator authority to the target

(especially when the target is protected), then sets status: applied.

  1. If rejecting or obsolete, sets status: closed.

Do not auto-apply AGENTS.md, context/, knowledge/, or other protected surfaces from a dream agent.

Agentic actions (reference)

KeyClassRole
self-orientagenticOrientation report + proposals
agentic-housekeepingagenticTidy survey, report-and-propose
edges-updatelight shelliris edges update --tier 0 --json
edges-densifyexpensive shelliris edges update --tier 2 --json

Agentic loops write only under forge/, inbox/captures/, and Lucerna runtime. Web tools are disabled for maintenance dreams. Wall-timeout tree-kill applies.

Enablement

Dreams default off. File: .amore/lucerna/enable.json with dreamsEnabled: true (or an equivalent start-time OR). Absent or malformed means false. --force on lucerna dream-cycle overrides schedule only, never enablement, and will not run a roster-disabled chore.

Reading budgets and the roster

Charter lives under .amore/lucerna/. Read it. Do not edit it.

FileWhat it is
.amore/lucerna/budgets.jsonSpend caps (dailyActionCap, weeklyExpensiveCap, cycleCooldownMinutes, dailyTokenCeiling, dreamsReserveTokens, autoCommitCooldownMinutes). Precedence is argv > env > file > shipped.
.amore/lucerna/chores.jsonNarrowing chore roster. Fields per key: enabled, minIntervalHours only. Unlisted keys stay enabled.
.amore/lucerna/enable.jsondreamsEnabled / autoCommitLive.
instruments/lucerna/state.jsonLive counters and last cycle outcome. Display only.

Also readable: iris lucerna budgets and iris lucerna chores list|show.

You may not edit `budgets.json`, `chores.json`, or `enable.json`. Those files are operator intent. Changing them is the operator's job (b / c on the Lucerna tab, or iris lucerna budgets set / iris lucerna chores enable|disable). A resident that rewrites the roster or the caps is authoring the steward's mandate.

Dreams-off is not spend-off: drafting still spends unless the operator set LUCERNA_AUTO_COMMIT=0.

Related

  • instruments/lucerna/README.md - enablement, budgets, agentic section
  • instruments/lucerna/src/engine/dispatch-contract.md - wall-timeout and web-off
  • House AGENTS.md skills table

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/vincitamore/amore-build/somniator/badge)](https://verified-skill.com/skills/vincitamore/amore-build/somniator)

Something wrong with this skill or its page? Tell us.