Content Pipeline
Sola Merlino turns a brand's own source content into platform-native posts and publishes them from real, warmed, account-safe browser identities — grounded in source, gated by account aging.
The pipeline has one job: take real content a brand already owns, rewrite it to fit each social platform, and post it from a fleet of identities that look and behave like ordinary, established accounts. Nothing in this flow fabricates engagement or manipulates content for deception — the adaptation is editorial (fit the platform), and the safety is operational (consistent network identity and human cadence). Below is the end-to-end path, followed by where each stage lives in the product.
The four stages
-
Source — grounding in the Ponmelli DAM
Every post starts from real, approved source material in the Ponmelli DAM, a SQLite asset catalog of 621 assets (
/api/ponmelli/*data/ponmelli/assets.db). Assets move through a lifecycle ofdraft → staged → approved → deployed → archived/rejectedand only approved content is eligible to feed the writer. This is the grounding layer: the writer adapts what already exists rather than inventing claims, so output stays anchored to the brand's own catalog. -
Adapt — Universal Writer, social type
The approved source is handed to the Universal Writer (
/api/universal-writer/*/api/universal-writer) running its social content type. The writer is RAG-grounded in the DAM source, so a single prompt produces per-platform variants for twitter, linkedin, facebook, instagram, reddit, tiktok and gmb, each respecting that platform's character limits, style and hashtag conventions. Variants are scored on Hook Strength 40% + Platform Fit 35% + Engagement 25%, with brand-voice override, link-anchor injection, and learned style rules drawn from prior human edits. -
Distribute — runners pick an identity and post
The cascade-runner and syndication-runner take the finished variants and select an eligible SyndicationPersona for the target platform. Each runner launches a hardened browser through the Antidetect Engine (port 3500), drives the live social session, and publishes the post. A persona is only chosen when its status is active and its aging stage permits the action — no identity posts before it is ready.
cascade-runner · syndication-runner -
Account safety — consistent identity, human cadence
Every persona carries a stable fingerprint and an assigned proxy region, bound together for a coherent network identity across sessions. Posting follows human cadence rather than burst automation, and actions are aging-gated by the persona's warming stage (0–7). Safety here is about the network footprint of the identity, not about altering the content it publishes.
/api/antidetect · /api/personas
Sola Merlino keeps accounts safe through IP and network hygiene — a consistent fingerprint, a stable proxy region, and human-paced behavior — so that a brand syndicating its own content reads as a normal, established account rather than as automation. This is legitimate business syndication. The platform does not manufacture engagement, falsify content, or provide operational guidance for deceiving platforms; adaptation is editorial fit, and protection is network-level identity consistency.
Where each part lives
Each stage of the pipeline maps to a dedicated area of the product. Follow these for the full detail on configuration, APIs and operation.
Adapt
The Universal Writer's social content type, RAG grounding, per-platform variants and scoring.
writer.htmlIdentities
Gmail accounts, brand workspaces, personas, warming, and the account lifecycle that gates posting.
accounts.htmlAccount safety
The Antidetect Engine: fingerprint isolation, behavioral entropy, and profile-to-proxy binding.
antidetect.htmlSource assets are managed in the Ponmelli DAM; approved content is what the writer is allowed to ground against, closing the loop from catalog to published post.
Sola Merlino · social syndication + account automation · documentation generated 2026-06-27