An organization's files hold its knowledge but do not surface it. An email archive, a folder of contracts, a store of PDFs — each is searchable, and none of it knows who relates to whom, which concerns recur, or what the organization's own terms mean.

service-content — the Gravity Engine — is the synthesis engine of the PointSav family. It reads raw payloads from inside a Totebox, runs them against an institutional taxonomy, produces dense Gravity Vectors, and generates the documents an organisation publishes: wikis, news releases, due-diligence ledgers, internal memos.

The engine is built on a Four-Pillar Seed Vault — the institutional grammar — a deterministic ingest-to-routing pipeline, and a five-layer Stratified Ontology. It is the difference between a file store and an active intelligence engine. The Chart of Accounts and eleven archetypes form two of the four Seed Vault pillars.

For a regulated buyer, one boundary governs all of it. service-content does not publish to a verified ledger autonomously; every verified entry transits a human-in-the-loop verification step first. This article covers the Seed Vault, the Gravity Engine pipeline, the Stratified Ontology, and that boundary.

The Four-Pillar Seed Vault

Every Totebox is provisioned with four JSON ledgers that form the institutional grammar of the system — the Seed Vault. The Seed Vault is built once, locked, and adjusted only by operators, never by AI.

A fifth pillar, Themes, captures the recurring topics and institutional concerns emerging from the corpus — flagged and proposed to operators for approval.

The Gravity Engine pipeline

When a payload arrives — an email, a PDF, a DOCX — the engine runs a tight, deterministic sequence.

1. Ingest. service-email, or the Input Machine, writes the raw payload to the Totebox's WORM vault. Nothing is mutated yet.
2. Entity extraction. service-extraction runs Aho-Corasick string matching ¹ over the raw text and pulls out every name, email, phone number, and organisation. Each receives a deterministic Sovereign-ID and begins in Discovery status.
3. Gravity calculation. service-content loads the four pillars, fuses them into a single search automaton, and extracts only the sentences containing structural keywords. A 5 MB payload becomes a 50-word Gravity Vector.
4. Verification. The vector and the entity bundle pass to service-slm, which evaluates whether the payload aligns with the institution's taxonomy. The output is a single token — VALID or REJECT.
5. Routing. Verified payloads go to deep storage; rejected payloads go to quarantine. Verified entities have their Discovery status upgraded and are socketed to the Chart of Accounts.

The Stratified Ontology

service-content organises information in a five-layer stack, with three derivative engines on top. The L0 base layer feeds directly into the WORM ledger architecture provided by service-fs.

Self-healing

The Gravity Engine is self-healing in a specific, narrow sense: its own outputs feed back into its inputs. A successfully synthesised memo becomes new content the engine indexes on its next run. Over months, the Domain glossaries grow, the Themes track real institutional concerns, and the engine's syntheses converge on the institution's actual voice. This is not autonomous model fine-tuning — the improvement is corpus-level: the engine has more verified material to draw from. The compounding substrate pattern describes the architectural logic behind this feedback loop.

The human-in-the-loop boundary

Every L5 entry must transit a human-in-the-loop verification step — through the Input Machine (F12) or the content review surface — before it can be written to a verified ledger. This satisfies SYS-ADR-07 (no structured data through AI) and SYS-ADR-19 (no automated AI publishing to verified ledgers).

The boundary is what makes the engine usable in a regulated setting. The engine accelerates synthesis; a human, not the engine, commits the result to a ledger of record.

Key takeaways

• service-content is the Gravity Engine — a synthesis engine that transforms an organisation's raw document payload into structured, human-verified intelligence.
• Its institutional grammar is the Four-Pillar Seed Vault: Entities, Archetypes, Chart of Accounts, and Domains — built once by operators and never mutated by AI.
• The pipeline is deterministic: Ingest → Extract → Gravity calculation → Verification → Routing. Every step produces auditable outputs.
• The Stratified Ontology has five layers. Only L5 — operator-verified content — feeds final published artefacts.
• The human-in-the-loop boundary is architectural, not optional: no content reaches a verified ledger without an explicit operator decision at the F12 gate.

See also

• service-slm — the local small language model that produces VALID/REJECT decisions
• service-people — the identity ledger that receives socketed entities from the Gravity Engine
• archetypes-and-chart-of-accounts — the institutional taxonomy that forms the Seed Vault's structure
• app-console-input — the F12 Input Machine; the human-gated ingest surface
• service-fs-architecture — the WORM ledger that stores the L0 base geometry layer
• totebox-os — the Totebox that hosts service-content and its WORM storage
• query-the-datagraph — step-by-step guide: search named entities, navigate relationships, and handle Tier A outages
• export-structured-data — step-by-step guide: four export paths including DataGraph queries and wiki Markdown