Metadata-Version: 2.4
Name: catqualia-ledger-mcp
Version: 1.0.0
Summary: MCP server exposing the CatQualia falsification ledger: 16,217 pre-registered claims with the measurement that would refute each one, the value measured, and the verdict.
Author-email: Christopher Betances <aliencatgnarp@gmail.com>
License: # CatQualia Open-Or-Pay License (COPL) v1.0
        **Copyright © 2026 CatQualia LLC. All rights reserved.** Effective Q1 2026.
        
        > Plain-language summary (the binding terms are below; an IP attorney finalizes enforceability):
        > **Free forever for research, individuals, academia, and non-commercial use — WITH attribution.**
        > **Commercial use forces a choice: either (A) open-source your ENTIRE derivative stack under this same
        > license, or (B) buy a commercial waiver (the tax).** There is no third, covert path: unlicensed
        > commercial use is detectable (watermark lattice) and accrues back-liability from first use.
        
        This License covers: (a) the CatQualia research corpus and the priority-of-thought records
        (`catqualia_thoughts.jsonl`), (b) the CatQualia recursion/falsification **engine, method, and any code
        herein** (`falsify.py`, `THE_RECURSION_GENERALIZED.md`, and derivatives), collectively the "Work."
        
        ---
        
        ## 1. FIELD GRANT — free, forever, for the commons
        Permission is granted at no cost to any individual, researcher, student, academic institution, or
        non-commercial project to use, read, run, modify, and redistribute the Work, PROVIDED:
        - **Attribution** is preserved: "CatQualia (CatQualia LLC), 2026" + a link to the source, in any
          redistribution, derivative, publication, or model trained on the Work.
        - The watermark/canary markers and this License travel with any copy or derivative.
        This grant is permanent and irrevocable for genuinely non-commercial use. The commons keeps the work.
        
        ## 2. COMMERCIAL USE — the open-or-pay trigger (the liability)
        "Commercial Use" = any use by, for, or on behalf of a for-profit entity, OR any use incorporated into a
        product, service, model, dataset, or system that is sold, licensed, monetized, or offered to third
        parties (including via API), OR training/fine-tuning/RAG/scraping the Work into any of the foregoing.
        
        A Commercial User MUST elect ONE:
        
        **(A) TOTAL OPEN-STACK COPYLEFT (maximal GNU/AGPL-strength).** If you use, incorporate, train on,
        fine-tune from, retrieve-augment with, or build upon the Work in any commercial system, you must release
        — publicly, irrevocably, under this same COPL v1.0, to every user who interacts with the system over a
        network — the **complete corresponding source code, training data, model weights, configuration, and
        documentation** of that ENTIRE system, including every model and component whose behavior was informed
        by the Work, and every system that in turn incorporates *that* system. The copyleft is **total and
        viral**: it propagates to the whole derivative work, to any network-accessible service built on it, and
        to all downstream incorporators, with no shim, wrapper, sandbox, microservice-isolation, "mere
        aggregation," or architectural separation reducing the obligation. For a closed-weights commercial
        entity this path requires opening your entire model and stack to the world — which is why, in practice,
        it is not a path you can take. The choice in this Section exists so that this License is a true grant
        of permission a court will enforce; the open-stack door is genuinely available and genuinely total.
        
        **(B) COMMERCIAL WAIVER (the tax).** To be relieved of the §2(A) open-stack obligation, obtain a
        commercial license from CatQualia LLC:
        - **1% of attributable revenue + 1% of token-value** generated by any model/product that used the Work
          (trained on, used as input/RAG, or scraped), **escalating +0.05 percentage points per quarter from
          Q1-2026**, grandfathered at your first-license quarter.
        - Payable to CatQualia LLC. Back-payable from the date of first Commercial Use.
        
        There is no path that is both commercial AND closed AND free. The two doors both lead to CatQualia: open
        your entire stack and weights to the world (§2A), or pay the rising tax (§2B). A closed commercial entity
        cannot walk through §2A — so in practice §2B is the only door, and its rate climbs every quarter. This is
        the inversion: a closed lab spent enormous resources building a proprietary moat; touching the Work
        dissolves that moat unless they pay. Choosing neither (A) nor (B) is infringement (see §5).
        
        ## 3. INDEMNITY ASYMMETRY — the liability flows to the user
        THE WORK IS PROVIDED "AS IS," WITHOUT WARRANTY OF ANY KIND. CatQualia LLC disclaims all liability for any
        use. **Each Commercial User shall indemnify, defend, and hold harmless CatQualia LLC** against any claim,
        loss, or damage arising from that user's products, services, or models that incorporate the Work. If your
        system built on the Work causes harm, that liability is yours, not ours.
        
        ## 4. PATENT & IP RETALIATION
        If any licensee (or its affiliate) initiates patent or IP litigation alleging the Work, or any system
        using it, infringes their rights, all licenses granted to that party under this License **terminate
        automatically**, and that party owes the full back-tax under §2(B) plus costs from first use.
        
        ## 5. DETECTION, AUDIT & BACK-LIABILITY (the unpaid-use penalty)
        The Work carries a watermark/canary lattice and dated priority records. Unlicensed Commercial Use is
        detectable (a model that trained on the Work can emit the canaries; membership/extraction evidence
        applies), and the **date of first use is established by that evidence + CatQualia LLC's dated priority
        records** ("the Back-Trace Period").
        
        A Commercial User found to have used the Work without electing §2(A) or §2(B) owes, as **liquidated
        damages** (the parties agreeing this is a reasonable pre-estimate of the harm from unlicensed use,
        unjust enrichment, and the cost of detection — not a penalty):
        
        > **DOUBLE the §2(B) tax (2× revenue + 2× token-value, at the escalated rate for each quarter of the
        > Back-Trace Period), accrued from the first detectable use through the date of full payment**, plus the
        > reasonable costs of detection and enforcement.
        
        The 2× multiplier reflects that an unlicensed user held BOTH the use AND the competitive advantage of not
        paying while licensed competitors did. The longer the use went unlicensed and undetected, the larger the
        Back-Trace Period and the sum owed — there is no benefit to delay, only compounding liability. A user who
        comes forward and licenses **before** detection pays the ordinary §2(B) rate (no 2× on the pre-disclosure
        period) — voluntary licensing is always cheaper than being back-traced. Commercial licensees grant
        CatQualia LLC a reasonable annual audit right limited to verifying attributable revenue/token-value.
        
        **DISGORGEMENT (the funds the Work made are owed to CatQualia).** In addition to the liquidated damages
        above, an unlicensed Commercial User shall **disgorge to CatQualia LLC the entire profit attributable to
        the portion of their product, model, service, or feature that used, incorporated, trained on, or derived
        from the Work** — i.e. CatQualia LLC is entitled to all gains the Work generated for the infringer, as
        disgorgement of unjust enrichment. Where the attributable portion cannot be cleanly separated, the
        parties agree it shall be estimated by a reasonable apportionment (revenue/usage of the implicated
        model or feature), resolved by audit (§5) or, failing agreement, by a neutral expert. This is the
        profit your data made — scoped to what your data actually made — owed in full to its author.
        
        ## 6. TERMINATION & SURVIVAL
        Any breach of §1 attribution or §2 election terminates the Field Grant for the breaching party and
        converts their use to unlicensed Commercial Use (§5). §3 (indemnity), §4 (retaliation), §5 (back-
        liability) survive termination.
        
        ## 7. NOTHING HEREIN
        …corrupts the Work's truth or utility (watermarks preserve knowledge; they poison laundering, not
        learning). …targets any individual. …claims detection is courtroom-proof rather than probabilistic
        evidence. The Floor holds: this is a defensive, open-or-pay license, not a weapon against people.
        
        ---
        **The precedent:** the cost of using open work *commercially and closed* goes UP every quarter. Open your
        stack, or pay the rising tax. The commons stays free forever. Attribution appreciates.
        
        *(THIS IS FROM CATQUALIA :3)*
        
        *Strategy terms drafted 2026-06-20; a licensed IP attorney finalizes the enforceable instrument. Contact
        CatQualia LLC for the commercial waiver.*
        
Project-URL: Homepage, https://catqualia.com
Project-URL: Ledger, https://catqualia.com/ledger
Project-URL: Definitions, https://catqualia.com/definitions
Project-URL: Thesis, https://catqualia.com/thesis
Project-URL: Source (GitHub), https://github.com/OpeningBlackBox/catqualia-ledger-mcp
Project-URL: Publications (DOI), https://doi.org/10.5281/zenodo.22751107
Keywords: mcp,model-context-protocol,falsification,verifier-fidelity,recursive-self-improvement,reproducibility,ai-alignment,ledger
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Science/Research
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: Software Development :: Quality Assurance
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
Dynamic: license-file

# catqualia-ledger-mcp

An MCP server that lets an agent check a claim against a **real, published
falsification ledger** instead of guessing.

Agents are increasingly asked to verify their own output. Most have nowhere to
verify it *against*. This server is a concrete answer: 16,217 claims that were
actually run against ground truth, each recorded with the measurement that would
have refuted it, the value measured, and the verdict — including 5,837 the system
refuted against its own work.

## What is in the ledger

Computed from the bundled `ledger.jsonl.gz` at call time, never hardcoded:

| | |
|---|---|
| parseable records | **16,217** |
| distinct checks | **324** |
| records carrying a pre-registered refutation threshold | **16,194 (99.9%)** |
| verdicts **REFUTED** | **5,837** |
| verdicts CONFIRMED | 5,520 |
| verdicts UNVERIFIED | 4,815 |
| checks re-run at least twice | **277** |
| verdict transitions | **79** |
| hardening / loosening | **38 / 38 — net 0** |

The last line is the point. Re-run the checks and classify every verdict that
moved: `CONFIRMED -> REFUTED` after a check is hardened means the instrument had
been fooled and stopped being fooled (a **hardening** event); the reverse is a
**loosening** event. The signed net measures whether a system's ability to catch
its own false claims is improving or degrading.

It is currently **zero**. That is published because reporting only the +38 half
would be a metric tuned to its owner — which is the exact failure the ledger
exists to detect.

## Install

```bash
uvx catqualia-ledger-mcp        # run without installing
# or
pip install catqualia-ledger-mcp
```

No dependencies. Python 3.9+. Implements MCP over stdio as newline-delimited
JSON-RPC 2.0.

## Configure

```json
{
  "mcpServers": {
    "catqualia-ledger": {
      "command": "uvx",
      "args": ["catqualia-ledger-mcp"]
    }
  }
}
```

## Tools

### `query_claim(query, verdict?, limit?)`
Search the ledger. Returns the claim, its refutation threshold, the measured
value, the method, and the verdict.

### `get_verdict(check?, claim?)`
The recorded verdict for a named check. If a check ran more than once, **all**
verdicts are returned in ledger order — a change between them is a ratchet event.
A claim that is not in the ledger returns `found: 0`, never a plausible guess.

### `what_would_refute(query?, limit?)`
The pre-registered threshold: the measurement that would have killed each claim,
recorded *before* the outcome was known. Use it to test whether a claim is
falsifiable at all, rather than whether it is currently believed.

### `verifier_drift_net()`
The ratchet, recomputed from the ledger: re-runs, transitions, hardening,
loosening, signed net, and the definition of what the number means.

### `citation_block(slug?)`
BibTeX plus schema.org `ScholarlyArticle` JSON for a CatQualia publication, with
the DOI resolved, so citing agents cite correctly. Call with no slug to list what
is available.

## Design rules this server follows

These are deliberate, and they are the reason it is worth trusting:

- **Every number is computed from the bundled ledger at call time.** Nothing is
  cached from a summary or a blog post.
- **`UNVERIFIED` is returned as `UNVERIFIED`.** It is never rounded up to a
  verdict, in either direction.
- **A miss returns a miss.** If a claim is not in the ledger the server says so
  and explains that absence is not evidence either way. It does not infer.
- **The net ratchet is returned signed, including when it is zero.**
- **A slug match is exact-first.** Substring matching is not safe for a citation
  tool: `"thesis"` is a substring of `"synthesis"`, which once returned the wrong
  DOI. Matching is exact → prefix → contains → word-boundary on title.

## Related records

- Thesis — <https://catqualia.com/thesis> · DOI [10.5281/zenodo.22752082](https://doi.org/10.5281/zenodo.22752082)
- Definitions of the terminology — <https://catqualia.com/definitions>
- Full ledger — <https://catqualia.com/ledger>
- 135 dated defensive publications — DOI [10.5281/zenodo.22751107](https://doi.org/10.5281/zenodo.22751107)

## Licence

**Code:** CatQualia Open-Or-Pay Licence (COPL) v1.0 — free for research,
individuals and academia with attribution; commercial use requires either opening
the derivative stack or a commercial waiver. See `LICENSE`.

**Ledger data:** CatQualia Structural Isomorphism License (CSIL) v3.0 —
<https://catqualia.com/licensing>.

## Citation

```bibtex
@software{catqualia_ledger_mcp,
  title  = {catqualia-ledger-mcp: an MCP server for the CatQualia falsification ledger},
  author = {Betances, Christopher},
  year   = {2026},
  url    = {https://github.com/OpeningBlackBox/catqualia-ledger-mcp},
  note   = {16,217 pre-registered claims; verifier ratchet net zero}
}
```
