Pre-ingestion verification for epistemic quality in RAG systems
Project description
clarity-gate
Pre-ingestion verification for epistemic quality in RAG systems.
Core Question: "If another LLM reads this document, will it mistake assumptions for facts?"
Status
⚠️ v0.1.0 is a placeholder release to reserve the package name. Full validation is coming soon.
For working validators now, use the existing packages:
pip install cgd-validatorpip install sot-validator
Or use the Claude skill directly.
What It Does
Clarity Gate validates documents before they enter knowledge bases, ensuring:
- Claims are properly marked as hypotheses vs. established facts
- Projections have uncertainty markers
- Assumptions are explicit, not buried
- Data is internally consistent
Installation
pip install clarity-gate
CLI Usage
# Validate a Clarity-Gated Document
clarity-gate validate-cgd document.cgd
# Validate a Source of Truth file
clarity-gate validate-sot reference.sot
# Auto-detect format
clarity-gate check document.md
# Options
clarity-gate validate-cgd document.cgd --json # JSON output
clarity-gate validate-cgd document.cgd --strict # Treat warnings as errors
Programmatic Usage
from clarity_gate import validate_cgd, validate_sot, validate
result = validate_cgd("path/to/document.cgd")
if result.valid:
print("Document passes all checks")
print(f"RAG-ingestable: {result.metadata.rag_ingestable}")
else:
for error in result.errors:
print(f"[{error.rule}] {error.message}")
Document Formats
CGD (Clarity-Gated Document)
Documents that have been verified for epistemic quality:
- YAML frontmatter with verification status
- Epistemic markers inline (e.g.,
*(not specified)*) - HITL verification records
- End marker:
Clarity Gate: CLEAR | REVIEWED
SOT (Source of Truth)
Authoritative reference documents:
- Strict structure requirements
- Staleness tracking with markers:
[STABLE],[CHECK],[VOLATILE] - Verified claims with citations
Specification
This validator implements:
- CGD_FORMAT.md v1.2 — 24 CGD rules
- SOT_FORMAT.md v1.2 — 7 SOT rules
- VALIDATOR_REFERENCE.md v1.2 — Implementation guide
Related Packages
| Package | Purpose |
|---|---|
clarity-gate |
Unified validator (this package) |
claritygate |
Alias → redirects here |
cgd-validator |
Legacy CGD validator |
sot-validator |
Legacy SOT validator |
Links
- clarity-gate — Main project repository
- arxiparse.org — Live implementation for scientific papers
- LessWrong post — Research writeup
License
CC-BY-4.0
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file clarity_gate-1.0.0.tar.gz.
File metadata
- Download URL: clarity_gate-1.0.0.tar.gz
- Upload date:
- Size: 19.1 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.13.2
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
ae0a30181cd47b4b9af5c5ece771c61dac8689e28c0fc6934440387eeaceed70
|
|
| MD5 |
78009b031387d5eb948a9b0c22c42ba0
|
|
| BLAKE2b-256 |
155986d07deb3bed8a65d32da730e6649926ece63ab89e1e8f868d7632101554
|
File details
Details for the file clarity_gate-1.0.0-py3-none-any.whl.
File metadata
- Download URL: clarity_gate-1.0.0-py3-none-any.whl
- Upload date:
- Size: 25.0 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.13.2
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
90bc2826a7dbe0a4889a3476befd6ca19d1836072a4d251528c0202db7261691
|
|
| MD5 |
31673c349b615c7666dc73dca98aa5fe
|
|
| BLAKE2b-256 |
5c91694536dcd61ee24eb0b1a28e766eaa50bbc2a4fdd4a45fab08e10323cdfb
|