fabius-archivumlisted

Maintain and retrieve permissioned project memory in the workspace's declared record store. Read a named project's record before work; preserve decisions, provenance, indexes, and history. Use for cross-session continuation, recording verified lessons, knowledge-base maintenance, legacy memory migration, meeting or video source ingest, and source-grounded notebook questions. One canonical store, gated writes, and fresh evidence before recalled conclusions on incident routes. Provider changes do not require moving records when the store is already shared.
shear559/fabius · ★ 2 · AI & Automation · score 64

Install: claude install-skill shear559/fabius

# Fabius Archivum — don't re-derive what you already learned *Archivum* — the record office, where what's settled is kept. An agent that re-researches every session pays full price each time. This layer makes knowledge compound: write it down, link it, retrieve it, keep it true. ## Three layers - **Raw sources** — the immutable source-of-truth documents. Read them, never edit them. - **The wiki** — an agent-owned directory of markdown pages (entities, concepts, comparisons, syntheses). This is the externalized long-term memory — the part that compounds. - **The schema** — a concrete config encoding the directory structure, the conventions, and the workflows. It turns a chat assistant into a disciplined archivist. Two navigation files keep hundreds of pages tractable: - **Index** — a catalog: every page with its link, a one-line summary, and metadata. Read this FIRST at query time, before opening any page. - **Log** — append-only, consistently prefixed, chronological. Greppable and tailable with plain unix tools. ## The three operations **Write boundary.** Archivum never turns a read-only task into a mutation. Ingest, capture, file-back, scaffold, and lint writes run only when the workspace has opted into Archivum memory **and** its contract authorizes those writes. Otherwise return a proposed record or diff without touchin