What we ship

Decipon scores influence tactics in articles, posts, and statements across twenty categories: emotional manipulation, missing information, tribal division, suspicious timing, uniform messaging, and the rest. Scores range from 0 (minimal tactics) to 100 (heavy use). It does not tell you what is true. It shows you how the content tries to influence you.

Every finding comes with the textual evidence that triggered it. Every score is paired with two perspectives: a critical read that weighs the manipulation signals, and a supportive read that weighs the underlying claims against available data. You see where they agree, where they diverge, and what to investigate further. The score is the start of the question, not the answer.

Built on the Influence Tactics Protocol

Decipon runs on the Influence Tactics Protocol (ITP), an open standard Synapti maintains. The methodology is auditable. Anyone can replicate the scoring, contest a finding, or extend the framework. Closed verdict tools that ask you to trust them are not the move.

Who uses it

Readers verifying news before they share it. Journalists running source material through a structured analysis pass. Researchers analyzing manipulation patterns at scale via API. Publishers integrating credibility scoring into editorial workflows. Different audiences, same principle: see the tactics, decide for yourself.

Where it lives

• Web app at decipon.com for analyzing pasted text or live news.
• Browser extension for inline scoring on the pages you read.
• API for batch analysis, dataset work, and editorial integration.
• Publisher badges that surface credibility scoring on the source itself.

Memomee is the memory governance and retrieval control plane for long-running AI agents. It keeps task state, decisions, and evolving facts compact, trustworthy, and resumable across sessions.

Most agents do not fail loudly. They fail by losing the plot: a session ends, a fact changes, or a summary compresses away the decision that mattered. Memomee gives that state an explicit lifecycle instead of asking a context window to hold it.

How it works

Agents write events and checkpoints as they work. Rules-based compaction merges repetitive events and flags contradictions. Retrieval returns a task-scoped bundle with the active checkpoint, filtered events, versioned facts, warnings, and an explanation of what was returned and why.

Who uses it

Teams building coding agents, research agents, enterprise workflows, support systems, and multi-agent handoffs that must survive longer than one session.

What it governs

• Checkpointed working state so an interrupted workflow can resume from the last successful step.
• Append-only events with actor, type, source, and payload provenance.
• Versioned facts that are superseded or retracted rather than silently overwritten.
• Explainable retrieval that reports what was returned, what was filtered, and why.

Loreon is an AI writing and novel-continuity platform for long-form and series fiction. It combines a structured story bible with scene, chapter, and whole-manuscript consistency checking.

It keeps characters, world rules, timelines, and story history as structured canon, then reads each passage against that canon. Canon links appear as the writer types, and consistency findings land in the margin only when asked.

Who uses it

Writers of long-form and series fiction who need to track large casts, layered timelines, world rules, and promises planted across chapters or volumes.

What it checks

• Structured canon for characters, locations, cultures, world rules, and story history.
• Scene and chapter continuity against the story bible.
• Whole-manuscript contradictions, timeline errors, voice drift, and dropped foreshadowing.
• Cross-volume continuity for series and sagas.

Specialist agents can draft or review, but they never touch the manuscript without the writer's say. The editor is free. Agent work runs on prepaid credits, not a subscription.