# Faultprint Source Freshness and Counter-Signal Pass 001

## Purpose

This pass adds two hardening layers to the 680-row crosswalk: source freshness metadata and counter-signal candidates. The goal is to make later benchmark and diagnostic claims more falsifiable.

## Source Freshness

- Unique URLs checked: 347
- fetch_supported_date_unknown: 169
- fresh_under_1_year: 131
- unverified_fetch_failed: 33
- usable_1_to_3_years: 9
- stale_over_3_years: 5

## HTTP Fetch Result

- 200: 314
- 403: 22
- 404: 7
- no_status: 2
- 500: 2

## Counter-Signal Candidates

- Candidate rows: 336
- mitigating_or_control_language: 206
- low_resemblance: 156
- hold_internal: 154
- absence_or_null_language: 144
- revise_internal: 31

## Pathology-Level Hardening

|pathology|rows|elevated/high|counter candidates|counter rate|dated sources|stale sources|freshness gap|
|---|---|---|---|---|---|---|---|
|Contestability Lag|249|82|206|0.8273|113|2|0.5462|
|AI Authority Acceleration|210|11|42|0.2|81|1|0.6143|
|Record-Authority Capture|70|38|23|0.3286|15|0|0.7857|
|Access Interruption Before Review|44|36|12|0.2727|24|0|0.4545|
|Classification Gatekeeping|27|0|10|0.3704|13|0|0.5185|
|Documentation Burden Transfer|26|16|11|0.4231|11|2|0.5769|
|Provisional Relief Reversal|24|0|24|1.0|17|0|0.2917|
|Evidence-Container Illusion|12|7|2|0.1667|3|0|0.75|
|Multi-Actor Handoff Diffusion|10|0|3|0.3|5|1|0.5|
|Deadline Compression|8|0|3|0.375|6|0|0.25|

## Plain-English Finding

The crosswalk has enough source anchoring for bounded assessment findings, but not enough source-date coverage for mature benchmark claims. The counter-signal layer is substantial and useful: it gives the system a way to say when a document contains controls, mitigants, absence statements, or low-resemblance findings instead of only surfacing exposure-like findings.

## What This Enables

1. A diagnostic report can include both findings and counter-signals.
2. A pathology module can show what would weaken or rebut the finding.
3. Benchmark claims can be gated until source freshness and sampling balance are adequate.

## Next Hardening Step

For the top five pathology modules, manually review the counter-signal candidates and promote true counterexamples into a curated counterexample library. Do not treat every low or held row as a true counterexample; many are only counter-signal candidates.