EditCred
EditCred
How it Works

Overview

EditCred is a human-in-the-loop provenance engine for AI-assisted writing. Currently browser-based.
It captures how professionals transform AI-generated drafts into structured, accountable final documents.

The Idea

AI is a powerful drafting partner.

But most professionals don’t publish raw AI output.

They:

  • Extract useful fragments
  • Rewrite in their own voice
  • Inject private context
  • Verify claims
  • Remove weak or risky phrasing
  • Combine multiple conversations into one coherent artifact
  • Align to their specific end-goals

That transformation is where the real work happens.

EditCred exists to recognize, structure, and validate that human effort.

What Happens Today

When you use AI for serious work, the process usually looks like this:

  1. You brainstorm with the model.
  2. You generate outlines or variations.
  3. You copy only what is useful.
  4. You restructure everything.
  5. You rewrite large portions.
  6. You add internal knowledge and constraints.
  7. You publish the final version aligned to your goal

The model generates drafts.
You generate judgment, coherence, and responsibility.

Until now, that transformation has been invisible.

How EditCred Works

Step 1 --- Use AI Normally

Work in ChatGPT or Gemini (for now).
No change in your creative process.

Editcred recognises prompt quality.

Step 2 --- Extract Intentionally

Highlight only the fragments that matter.
Discard filler.
Combine pieces from multiple responses.

EditCred captures this extraction layer.

Step 3 --- Structure in the Side Panel

Selected content appears in a structured editor in your browser.

You can:

  • Rearrange sections
  • Create headings
  • Merge fragments
  • Organize your document logically

This is where structure replaces raw output.

Step 4 --- Transform

You refine the draft:

  • Rewrite sentences
  • Add private data and insights
  • Correct inaccuracies
  • Adjust tone
  • Remove hallucinations
  • Compress or expand sections
  • Synthesize multiple threads

EditCred captures how your document evolves from draft to final artifact.

Step 5 --- Human Agency Score

EditCred analyzes the difference between:

  • The AI’s draft
  • Your final document

From this, a Human Agency Score is calculated.

This score reflects transformation --- not just prompting.

It recognizes:

  • Goal alignment
  • Structural changes
  • Degree of rewriting
  • Context injection
  • Multi-source synthesis
  • Iterative refinement

The goal is not to reduce AI usage.
It is to recognize meaningful human contribution and efficiency.

Join our community.

Why This Community Exists

We believe:

  • AI is a drafting tool, not a replacement for judgement.
  • Editing is the real skill.
  • Professional credibility requires accountability.
  • Human contribution in AI-assisted writing should be measurable and verifiable.

This community is for people who:

  • Care about the quality of their output
  • Publish public-facing work
  • Operate in high-stakes environments
  • Use AI thoughtfully and responsibly

What’s Coming

We are building:

  • Verifiable provenance support (C2PA-compatible manifests)
  • Public authenticity signals
  • Aggregated insights into real-world AI editing patterns
  • Tools to improve LLM performance using post-edit data

Everything is opt-in.
You control what is shared.

Founding Community

EditCred is currently in controlled beta.

Members of this community:

  • Get early access
  • Influence product direction
  • Shape how human-in-the-loop AI work is measured
  • Help define standards for responsible AI-assisted writing

If you believe AI drafts are just the beginning --- and the real work is in the edit --- you belong here.