SIC[]

AI memory that tells your story, as written.

Deck AI generated with the aswritten.ai storyBASE

The Problem

AI without memory is generic

What is storyBASE?

Git-native RDF knowledge graph

Your narrative source of truth

The Mission

Extend software development rigor into strategy, content, and marketing. Provide versionable, collaborative, narrative-driven AI memory.1

  1. Mission statement from http://storybase.synthetic-identity.co/mission/storybase. This positions storyBASE as bringing Git-like version control and collaboration patterns to organizational narrative and strategy work. ↩︎

How It Works

  1. Extract narrative from inputs (transcripts, docs, conversations)
  2. Diff proposed changes against current snapshot
  3. Commit transactions to append-only log
  4. Compile snapshot from sorted transaction history
  5. Generate stories using snapshot as context
Architecture

Immutable by design

Append-Only Transaction Log

Every change is a transaction. Transactions are immutable files. The snapshot is the replay of sorted transactions. Provenance is built in.1

  1. Data Model Lifecycle (http://storybase.synthetic-identity.co/model/data-lifecycle-storybase) specifies “Append-only transaction log; immutable files; snapshot = replay of sorted transactions; provenance in TX step; future named graphs for add/remove.” ↩︎

System Topology

n8n agent orchestrates tools

MCP server exposes to frontends (Agent.ai, ChatGPT, Open WebUI)

Transactions in /.storybase directories

Hierarchical compile from root to leaves

Docker Compose on Digital Ocean1

  1. System Topology (http://storybase.synthetic-identity.co/architecture/topology-storybase) describes the orchestration layer and integration points. The MCP (Model Context Protocol) server provides a standard interface for AI frontends to access storyBASE tools. ↩︎

Integration Points

GitHub: OAuth, webhooks, Actions

Open Router: API proxy via Helicone

Outseta: OIDC, billing

MCP protocol: tool exposure

Future: GitHub Apps with scoped credentials1

  1. Integration Points (http://storybase.synthetic-identity.co/integration/points-storybase) and Dependencies (http://storybase.synthetic-identity.co/dependency/storybase-integrations) detail the external systems. Helicone provides API monitoring; Open Router proxies model access; Outseta handles authentication and billing. ↩︎

Narrative Architecture

The operating system for story-led strategy

Six Core Domains

  1. Opportunity: Market context, actors, timing
  2. Strategy: Positioning, moat, narrative anchor
  3. Product: Capabilities, flows, offerings
  4. Architecture: System design, explainers, docs
  5. Organization: Roles, processes, change management
  6. Proof: Case studies, outcomes, metrics1

  1. The Narrative Architecture ontology defines these six top concepts (http://example.org/narrative#Opportunity, #Strategy, #Product, #Architecture, #Organization, #Proof) as the foundational structure. This framework appears in the ontology provided in the SNAPSHOT. ↩︎

Style as Structure

Diction: Terminology control, naming conventions, verb choice

Tone: Direct/personal, authoritative, active voice

Cadence: Sentence length variation, rhythm, rule of three

Devices: Simile, metaphor, analogy, rhetorical questions

Metrics: Readability, active voice ratio, jargon density1

  1. The Style top concept (http://example.org/narrative#Style) was added to the ontology to encode linguistic and presentation features. It includes facets for diction, tone, grammar, cadence, rhetorical devices, orthography, punctuation, citation conventions, and more. ↩︎

Conviction Levels

Notion: Exploratory, open graph edges

Stake: Proposed, has supporting value

Boulder: Settled, hard to move

Foundation: Underpinning, effectively permanent1

  1. The Conviction top concept (http://example.org/narrative#Conviction) encodes “degree of settledness of a claim, from loose notions to foundations; used to govern decisions and change cost.” The four levels use xkos:next/previous to encode escalation paths. ↩︎

Positioning

Where we play, how we win

Primary Actors

Programming-literate entrepreneurs, designers, developers, consultants who manipulate worldview and see perspective changes.1

  1. Primary Actors (http://storybase.synthetic-identity.co/actor/primary-storybase) defines the target: “Programming-literate entrepreneurs, designers, developers, consultants who manipulate worldview and see perspective changes.” ↩︎

Moat & Leverage

Git-native, versionable, branchable AI memory encoding style, conviction, narrative metrics. Replaces brittle role prompts with deep, operable persona descriptions.1

  1. Moat Leverage (http://storybase.synthetic-identity.co/leverage/moat-storybase) describes the defensibility: “Git-native, versionable, branchable AI memory encoding style, conviction, narrative metrics; replaces brittle role prompts with deep, operable persona descriptions.” ↩︎

Timing Thesis

Convergence of prompt engineering maturity, multi-agent workflows, and demand for organizational AI memory creates window for narrative-driven context management.1

  1. Timing Thesis (http://storybase.synthetic-identity.co/thesis/timing-storybase) identifies the convergence: “Convergence of prompt engineering maturity, multi-agent workflows, and demand for organizational AI memory creates window for narrative-driven context management.” ↩︎

Proof

It works

This deck was generated from a storyBASE snapshot and a .story template

Rubric Assessments

Strategic Alignment: 4.0/5 — Clear positioning, mission, moat

Technical Depth: 4.8/5 — Grounded in Clojure principles, verifiable architecture

Narrative Coherence: 4.6/5 — Coherent arc from problem to proof

Accuracy: 4.0/5 — Technical details specific, named entities correct1

  1. Rubric assessments from the Conj 2025 extraction (urn:uuid:rubric-clarity, urn:uuid:rubric-technical-depth, urn:uuid:rubric-narrative-coherence, urn:uuid:rubric-accuracy) provide quantitative validation of narrative quality across multiple dimensions. ↩︎

Style Metrics

Average sentence length: 22.4 (Conj talk), 28.5 (Sample 1), 35.2 (Check-in)

Active voice ratio: 0.71–0.75

Jargon density: 0.12–0.18

Technical density: 0.68 (Conj talk)1

  1. Style Metrics from three samples: Conj Talk (urn:uuid:style-metrics), Sample 1 (http://example.org/narrative#Metrics_Sample1), and Check-in (http://storybase.synthetic-identity.co/metrics/style) show variation across contexts while maintaining recognizable patterns. ↩︎

Transactions as Provenance

Every claim in this presentation traces back to a transaction:

Tx_20251110T184512Z_sample1: Narrative architecture extraction

2025–01–29T000000Z_sic-storybase-checkin: Product/strategy check-in

Tx_20251109T223928Z_conj2025: Conj 2025 talk extraction1

  1. The three transactions (http://example.org/narrative#Tx_20251110T184512Z_sample1, http://storybase.synthetic-identity.co/transaction/2025–01–29T000000Z_sic-storybase-checkin, http://example.org/narrative#Tx_20251109T223928Z_conj2025) establish the current state of the storyBASE graph. ↩︎

The Vision

AI that tells your story, as written

Tagline

AI that tells your story as written1

As written carries double meaning: i.e. (that is to say) and as you wrote it. The brand is aswritten.ai.2

  1. Tagline (http://storybase.synthetic-identity.co/tagline/storybase) with note: “User-facing brand as written.ai; Latin i.e. meaning.” The tagline encodes both the product promise (AI that respects your narrative) and the brand identity. ↩︎

  2. The Latin i.e. (id est, “that is”) connects to the brand name “Sic” (Latin for “thus” or “as written”), creating a coherent linguistic identity around fidelity to source material. ↩︎

What’s Next

Build the marketplace: Shareable narrative modules

Prove the model: Design partner beta, public case studies

Extend the ontology: More domains, deeper conviction modeling

Scale the community: Programming-literate strategists who see the pattern1

  1. The roadmap (http://storybase.synthetic-identity.co/roadmap/narrative-storybase) and case studies (http://storybase.synthetic-identity.co/case/studies-storybase) outline the path from prototype to platform. ↩︎

Compile your org.

contact for demo:

scarlet@aswritten.ai