Home / yogsoth-ai / de-anthropocentric-research-engine · skills/artifact-detection/SKILL.md · GitHub

artifact-detection skillA

artifact-detection is agent-read markdown (skill) from yogsoth-ai/de-anthropocentric-research-engine: Detect annotation artifacts and shortcuts in benchmarks.

Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.

What the file says

# Artifact Detection Tactic

Systematically probe benchmarks for annotation artifacts, dataset shortcuts, and spurious correlations that allow models to achieve high scores without the intended capability.

## Stages

### Stage 1: Hypothesis-Only Baseline Test

Search literature for evidence that partial-input baselines achieve unexpectedly high performance:
- Hypothesis-only baselines (NLI without premise)
- Question-only baselines (QA without context)
- Label-word frequency baselines
- Majority-class and surface-pattern baselines

**Search queries**: "[benchmark] annotation artifacts", "[benchmark] hypothesis only", "[benchmark] spurious correlations", "[benchmark] dataset bias"

If published partial-input results exist, record performance gap between partial and full input. Gap < 10 points above random indicates severe artifacts.

### Stage 2: Contrast Set Construction

Identify whether contrast sets or adversarial evaluations exist:
- Search for "[benchmark] contrast sets", "[benchmark] adversarial examples"
- Check if CheckList-style behavioral tests have been applied
- Look for counterfactual data augmentation studies
…

Read the whole file at its exact version.

How to install

Latest version
mdr add yogsoth-ai/de-anthropocentric-research-engine/artifact-detection@git:20260615.7a62936
Exact content
mdr add yogsoth-ai/de-anthropocentric-research-engine/artifact-detection@sha256:a4fc474451467404

Pin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_7szej6pj2lcdxpfi.svg)](https://markdownregistry.com/a/art_7szej6pj2lcdxpfi)

1 badge views in 30 days

Versions

versioncommittedcommitsizeaudit
git:20260615.7a62936 latest2026-06-15 7a62936 3,076 BA view · diff
git:20260519.dc15fe22026-05-19 dc15fe2 3,107 BA view

Audit of the latest version

A  17 of 17 checks passed. Deterministic, no model, same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (3076 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

yogsoth-ai/de-anthropocentric-research-engine · 498 stars · license Apache-2.0 · pushed 2026-09-17 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_7szej6pj2lcdxpfi
GET https://markdownregistry.com/api/v1/resolve?ref=yogsoth-ai/de-anthropocentric-research-engine/artifact-detection
GET https://markdownregistry.com/api/v1/blob/a4fc4744514674042cdbaafe211a4babf0f0fb79339ca9c0b700b6bc647d5453

Your agent does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

More from yogsoth-ai/de-anthropocentric-research-engine

AGENTS.md agents
yogsoth-ai/de-anthropocentric-research-engine · AGENTS.md
git:20260727.358b018 · audit A · 498 stars
CLAUDE.md@ladder-foundry claude
yogsoth-ai/de-anthropocentric-research-engine · ladder-foundry/CLAUDE.md
git:20260701.4f1b4d0 · audit A · 498 stars
formated-results skill
yogsoth-ai/de-anthropocentric-research-engine · ladder-foundry/skills/formated-results/SKILL.md · Closing skill for the research-executor, loaded as the last step of formated-specs. Summarize the design just produced…
git:20260916.f5399c9 · audit A · 498 stars
formated-specs skill
yogsoth-ai/de-anthropocentric-research-engine · ladder-foundry/skills/formated-specs/SKILL.md · Spec-slot skill for the research-executor. Emit the 4-layer DARE orchestration of the assigned topic as one…
git:20260916.f5399c9 · audit A · 498 stars
injection-fidelity skill
yogsoth-ai/de-anthropocentric-research-engine · ladder-foundry/skills/injection-fidelity/SKILL.md · Loss-1 judge (codex role). Given one sample's de-identified dialogue and its PolicyCard, decide axis-by-axis whether…
git:20260916.f5399c9 · audit A · 498 stars
ladder-quality-order skill
yogsoth-ai/de-anthropocentric-research-engine · ladder-foundry/skills/ladder-quality-order/SKILL.md · Loss-2 judge (codex role). Over one topic's 6 shuffled research-design samples, pairwise-rank by quality using the…
git:20260916.f5399c9 · audit A · 498 stars
optimization-loop skill
yogsoth-ai/de-anthropocentric-research-engine · ladder-foundry/skills/optimization-loop/SKILL.md · The optimizer brain for the ladder-foundry pretraining loop. Runs the two-level nested batch loop, delegates gating to…
git:20260916.f5399c9 · audit A · 498 stars
acu-nugget-recall skill
yogsoth-ai/de-anthropocentric-research-engine · paper-reading/skills/acu-nugget-recall/SKILL.md · Tactic: Extract atomic units from one paper and score how much of a caller-supplied summary covers. Use for ACU-style…
v1.0.0 · audit A · 498 stars
argumentative-zoning skill
yogsoth-ai/de-anthropocentric-research-engine · paper-reading/skills/argumentative-zoning/SKILL.md · Tactic: Label every sentence of one paper with its rhetorical role using Argumentative Zoning. Use when fixed…
v1.0.0 · audit A · 498 stars
atomic-unit-matching skill
yogsoth-ai/de-anthropocentric-research-engine · paper-reading/skills/atomic-unit-matching/SKILL.md · Judge, per atomic content unit, whether a target text (summary, abstract, or other candidate text) contains it — binary…
v1.0.0 · audit A · 498 stars
atomic-unit-recall-aggregate skill
yogsoth-ai/de-anthropocentric-research-engine · paper-reading/skills/atomic-unit-recall-aggregate/SKILL.md · Aggregate per-unit ACU/Nugget match judgments into a final recall score — normalized length-penalized recall for ACU…
v1.0.0 · audit A · 498 stars
atomic-unit-writing skill
yogsoth-ai/de-anthropocentric-research-engine · paper-reading/skills/atomic-unit-writing/SKILL.md · Extract (ACU-style) or freshly author (Nugget-style) a list of atomic content units from a paper, optionally tagged…
v1.0.0 · audit A · 498 stars

Every file in yogsoth-ai/de-anthropocentric-research-engine

Browse by kind, by grade A, or by owner.