> ## Documentation Index
> Fetch the complete documentation index at: https://mehen.ophi.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Filler / Lazy Structure Risk

> Structural evidence for 'big but vacuous' documents — without claiming anything about how the text was written.

This metric addresses the AI-era documentation problem:

> The document is just filler: structure is lazy, there are no references, it is large but useless.

<Warning>
  This is **not** AI-authorship detection. It reports structural evidence — unanchored prose, low artifact
  density, weak repository grounding, lazy sectioning, repetition, specificity scarcity, hollow
  references, and placeholder density — without making any claim about how the text was written.
</Warning>

## Sub-scores (.1–17.8)

| Sub-score               | What it captures                                                                                     |
| ----------------------- | ---------------------------------------------------------------------------------------------------- |
| **UnanchoredProseMass** | Fraction of words living in sections with no evidence anchors.                                       |
| **LowArtifactDensity**  | `1 − sat(A / (W/800); 0.5, 2.0)` — too few code, tables, diagrams.                                   |
| **LowRepoGrounding**    | `1 − RepositoryGroundingScore`.                                                                      |
| **LazySectioning**      | Heading density, large-section rate, "shallow big doc" flag (`W > 2,500` AND max heading depth ≤ 2). |
| **RepetitionDensity**   | Token-shingle Jaccard > 0.82 detects near-duplicate paragraphs.                                      |
| **SpecificityScarcity** | Identifiers + paths + version tokens + inline code tokens relative to `W`.                           |
| **ReferenceHollowness** | Bibliography entries without verifiable DOI/arXiv/RFC/URL anchors.                                   |
| **PlaceholderDensity**  | TODO/TBD/FIXME/XXX/lorem and empty links per 1,000 words.                                            |

## Formula

```text theme={null}
FillerLazyRisk = clamp01(
    0.20 · UnanchoredProseMass
  + 0.15 · LowArtifactDensity
  + 0.20 · LowRepoGrounding
  + 0.15 · LazySectioning
  + 0.12 · RepetitionDensity
  + 0.12 · SpecificityScarcity
  + 0.04 · ReferenceHollowness
  + 0.02 · PlaceholderDensity
)
```

## Bands

| Score       | Band    |
| ----------- | ------- |
| 0.00 – 0.20 | Low.    |
| 0.21 – 0.40 | Mild.   |
| 0.41 – 0.60 | Review. |
| 0.61 – 0.80 | High.   |
| 0.81 – 1.00 | Severe. |

## Diagnostic labels

High scores attach stable string labels reviewers can act on:

* `large-unanchored-prose`
* `low-repository-grounding`
* `lazy-sectioning`
* `low-artifact-density`
* `near-duplicate-paragraphs`
* `specificity-scarcity`
* `hollow-references`
* `placeholder-heavy`

The PR comment quotes these labels verbatim instead of paraphrasing.

## Example output

```text theme={null}
Filler / Lazy Structure Risk: 0.73 HIGH

Top contributors:
  - 71% of prose is in sections without evidence anchors
  - 3,420 words, only 1 relative link and 0 code examples
  - max heading depth = 2 with 4 sections > 1,200 words
  - specificity density = 1.8% (threshold: 3%-15%)
```

## References

* Pirolli, P. & Card, S. (1999). *Information Foraging.* Psychological Review 106(4): 643–675 —
  motivates the evidence-anchor and specificity-scarcity sub-scores.
  [DOI](https://doi.org/10.1037/0033-295X.106.4.643).
* Halliday, M. A. K. (1985). *Spoken and Written Language.* Oxford University Press — lexical-density
  basis used by `SpecificityScarcity`.
* Manning, C. D., Raghavan, P. & Schütze, H. (2008). *Introduction to Information Retrieval*, ch. 6.
  Cambridge University Press — Jaccard / token-shingle methods used by `RepetitionDensity`.
  [Stanford online edition](https://nlp.stanford.edu/IR-book/).

## See also

* [DMI](/metrics/markdown/dmi) — uses Filler Risk as one of its inputs.
* [Repository Grounding](/metrics/markdown/repository-grounding) — feeds LowRepoGrounding.
* [Evidence Coverage](/metrics/markdown/evidence-coverage) — feeds UnanchoredProseMass.