Stats
Actions
Tags
From vicky
Walk the KB and connect what is already there. Drains the pending queue into sources and rebuilds the graph. No external fetches, no stub conclusions — conclusions are only born from real synthesis via `conclude` / `complete-research`. For new external data use /vicky:research, which calls this skill at the end of its own pipeline. Use /vicky:learn on its own when pending notes arrived from elsewhere (git pull, sibling agents, cron) and just need absorbing.
How this skill is triggered — by the user, by Claude, or both
Slash command
/vicky:learnThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Drain pending → sources → relink. Pure DB walk; no web, no stubs.
Drain pending → sources → relink. Pure DB walk; no web, no stubs.
The transitions are enforced in code (learn MCP tool), not in this
document.
Invoke: /vicky:learn
For every file in vicky/pending/:
vicky/sources/<slug>.md with
type: source, tags: [source], body = question + context + graph
context. Any existing source links from the pending stub end up as
related: on the new source.related: frontmatter refreshed on every
note across sources and conclusions.No conclusion is created. Conclusions land in vicky/conclusions/
only when an agent has a real synthesis to write — via conclude
(direct) or complete-research (research/ promotion). The dashboard
"Sources awaiting synthesis" section surfaces sources without an
inbound conclusion so the work is visible.
Idempotent: if a source already exists for a pending slug, the pending file is dropped without overwrite.
WORKFLOW.md → default_workflow: triage filters the drain to
priority: high pending notes only. Other priorities stay queued.
git pull, or cron and you
want it absorbed before querying.enqueue calls outside /vicky:research.Do not call for an individual question the KB might already answer
(use research-gap). Do not call when you also need fresh external
data — use /vicky:research, which fetches and then invokes learn
automatically.
learn leaves new sources without paired conclusions on purpose. Open
Dashboard.md (or call dashboard) and look at "Sources awaiting
synthesis": those are the candidates for the next conclude pass.
Read the source body, distil the takeaway, then call conclude title=<slug> sources=[<slug>, ...].
vicky/pending/*.md deletedvicky/sources/*.md added (promoted from pending)vicky/graphs/*.json rebuiltvicky/sources/*.md related: updated by relinkvicky/conclusions/*.md related: updated by relink## Question — filename used as the question.Searches MemPalace before answering questions about past work, people, projects, or prior decisions. Returns verbatim stored content instead of guessing from model memory.
Guides Payload CMS config (payload.config.ts), collections, fields, hooks, access control, APIs. Debugs validation errors, security, relationships, queries, transactions, hook behavior.
Implements vector databases with Pinecone, Weaviate, Qdrant, Milvus, pgvector for semantic search, RAG, recommendations, and similarity systems. Optimizes embeddings, indexing, and hybrid search.
npx claudepluginhub yesitsfebreeze/vicky --plugin vicky