Live survey304 models · 14 providers · 23,592 auditsAudited implement · SAFE →What changedHow we verifyllms.txt

Agent skill · Audited from source

Research Catalog

skill · yogsoth-ai/de-anthropocentric-research-engineApache-2.0 licenceAudited 19 Oct 2026 · source e94c5ee067cd

★ 504pushed 1 day ago
✓SAFEGrade B · 89/100 · 19 Oct 2026 · e94c5ee
504repo stars · whole yogsoth-ai/de-anthropocentric-research-engine

Is it safe? Yes — SAFE, grade B (89/100). We read this skill's source at commit e94c5ee067cd and found no filesystem, network or shell access and no hard-coded secrets. What it does: This is the v4 capability menu. It indexes exactly 51 tactics in 10 families and tells the caller when each tactic is useful.

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “Research Catalog (yogsoth-ai/de-anthropocentric-research-engine)”. Retrieved 19 Oct 2026. https://modelsatlas.com/skills/yogsoth-ai/research-catalog

Trust audit

Audit 0.4.1 · source read 19 Oct 2026

Trust score89/100✓SAFE B
0 critical0 high0 medium0 low0 info

SAFE means nothing in the source contradicts what the skill says it does. Grade B is shared by 17,406 of 18,389 audited skills.

What the source does

  • FilesystemNone observedNo reads or writes in the source
  • NetworkNone observedNo outbound calls in the source
  • ShellNone observedNo commands run in the source
  • DependenciesPinnedNothing floats
  • SecretsNone foundNo keys in source

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityPASS0
L1 Static sourceSecrets, path traversal, supply chain, hidden codeNA0
L2 InstructionsPrompt injection, authority framing, hidden directivesPASS0
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: no behavioural pass.

Findings (0)

No findings at any severity. We re-audit when the source changes.

Audit 0.4.1 · source sha e94c5ee067cd · 19 Oct 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • Publisheryogsoth-ai GitHub organizationOwner of the repository this skill is read from
  • Source folderyogsoth-ai/de-anthropocentric-research-engine/skills/research-catalogThe exact folder we audited
  • LicenceApache-2.0As the repository states it
  • Commit pinned for this verdicte94c5ee067cdA new audit runs when the source changes
  • Repository statusNot archivedLast push 1 day ago
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: This is the v4 capability menu. It indexes exactly 51 tactics in 10 families and tells the caller when each tactic is useful.

---
name: research-catalog
description: "This is the v4 capability menu. It indexes exactly 51 tactics in 10 families and tells the caller when each tactic is useful."
---

# research-catalog

## Purpose

This is the v4 capability menu. It indexes exactly 51 tactics in 10 families and tells the caller when each tactic is useful. It does not list SOPs. A selected tactic owns its SOP calls through `You MUST load skill ...` instructions in the tactic body.

## Input Contract

```yaml
required: [confirmed_north_star, research_brief]
optional: [current_spec_view, state_slice, constraints]
constraints:
  - return tactic cards, not SOP cards
  - preserve the tactic id exactly as listed
```

## Execution Protocol

Read the North Star, ResearchBrief, current SpecView, and relevant state slice. Select 3-5 candidate tactics whose purpose and prerequisites match the current objective. Return their family, id, why-now condition, required inputs, expected outputs, and next call. Do not invent an order for tactics that the Spec has not ordered. Once a tactic is selected, the caller MUST load that tactic skill; its body is authoritative for SOP calls and thresholds.

## Tactic Index

### ACQUISITION

| Tactic | When to use |
| --- | --- |
| `synthesize-literature-evidence` | When the question needs a structured synthesis of literature claims and evidence. |
| `mine-patent-landscape` | When patents, assignees, classifications, or filing trends define the search space. |
| `assess-prior-art-and-claims` | When a proposed contribution must be tested against prior art and claim boundaries. |
| `map-patent-white-space` | When patent coverage and uncrowded opportunity regions must be compared. |
| `audit-benchmark-validity` | When benchmark metrics, datasets, or comparison protocols may distort conclusions. |
| `synthesize-meta-analytic-evidence` | When multiple studies provide comparable effect estimates requiring pooled interpretation. |
| `establish-empirical-baseline` | When a new method needs a reproducible baseline and a defensible comparison reference. |

### CONVERGENCE

| Tactic | When to use |
| --- | --- |
| `pairwise-ranking` | When candidates are easier to compare two at a time than on a shared absolute scale. |
| `structured-consensus` | When several judgments must be reconciled with explicit disagreement and evidence. |
| `portfolio-optimization` | When the result should be a balanced subset rather than one winner. |

### CROSS

| Tactic | When to use |
| --- | --- |
| `rank-candidates` | When typed candidates must be scored, screened, classified, or selected under stated criteria. |
| `map-validity-envelope` | When the conditions under which a claim or design remains valid are unknown. |
| `explore-dimensional-space` | When a design space needs axes, combinations, and coverage gaps. |
| `adversarial-deliberation` | When a consequential decision needs structured opposing cases before commitment. |
| `analyze-constraints-readiness` | When constraints, dependencies, conflicts, and readiness determine whether work can proceed. |

### DIRECTION

| Tactic | When to use |
| --- | --- |
| `map-research-landscape` | When fields or subfields must be compared by maturity, competition, barriers, and opportunity. |
| `decompose-research-goal` | When a broad goal must become ordered, testable objectives and dependencies. |

### EXPERIMENT

| Tactic | When to use |
| --- | --- |
| `design-experiment` | When hypotheses need factors, variables, controls, measurements, and analysis plans. |
| `analyze-future-scenarios` | When a design or strategy must be tested against plausible future conditions. |
| `analyze-experiment-results` | When observations must be interpreted against the preregistered design and gates. |

### HYPOTHESIS

| Tactic | When to use |
| --- | --- |
| `formulate-hypotheses` | When observations or gaps must become structured, named, testable hypotheses. |
| `falsifiability-audit` | When a hypothesis needs explicit disconfirmation conditions and observable consequences. |
| `formulate-research-question` | When an intent must become a precise, bounded research question. |
| `decompose-research-question` | When one question contains separable subquestions with distinct evidence needs. |

### IDEATION

| Tactic | When to use |
| --- | --- |
| `analogical-discovery` | When another domain may provide a transferable mechanism or design pattern. |
| `destructive-ideation` | When assumptions should be inverted or deliberately broken to expose alternatives. |
| `structural-transformation` | When an existing concept should be decomposed and recombined structurally. |
| `coverage-white-space-search` | When the candidate space is incomplete and missing regions need systematic search. |
| `resolve-inventive-contradiction` | When requirements conflict and a solution must preserve both sides. |
| `biomimetic-transfer` | When biological functions or mechanisms may transfer to the target problem. |
| `conceptual-blending` | When two or more concept spaces should be combined to produce emergent candidates. |
| `evolve-solution-population` | When iterative variation and selection are more useful than one-shot ideation. |

### INSIGHT

| Tactic | When to use |
| --- | --- |
| `validate-research-gap` | When an alleged gap needs evidence, novelty, and boundary verification. |
| `drill-root-causes` | When surface symptoms must be traced to mechanisms and contributing conditions. |
| `assumption-stress-test` | When hidden assumptions may determine whether a conclusion survives. |
| `robustness-analysis` | When conclusions must be checked across perturbations, alternatives, or distributions. |
| `sensitivity-analysis` | When priority depends on which inputs or assumptions drive the result. |
| `problem-reframing` | When the current problem statement may encode an unhelpful framing or boundary. |
| `map-stakeholder-system` | When actors, incentives, dependencies, and tensions shape the research problem. |

### STRESS

|

Excerpt, first 6,000 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit e94c5ee

Download folder ↗
git clone https://github.com/yogsoth-ai/de-anthropocentric-research-engine
git -C de-anthropocentric-research-engine checkout e94c5ee067cd

The author documents no install command, so this uses the repository directly. Our verdict covers commit e94c5ee067cd only.

Works in

The author has not said which agents this skill supports, and we have not tested it in any.

Audit history

Every audit, every source change

  1. ✓SAFE B 89 / 100

    first audit · source e94c5ee067cd

Broad Web Search✓SAFE

yogsoth-ai/de-anthropocentric-research-engine

A 267-skill research graph in pure markdown — 51 research operations built from 216 single-purpose steps, composed in any order with explicit backtracking. One npx install, no runtime, no MCP bindings. The AI is the…

★ 504Grade B
Docs✓SAFE

yogsoth-ai/de-anthropocentric-research-engine

A 267-skill research graph in pure markdown — 51 research operations built from 216 single-purpose steps, composed in any order with explicit backtracking. One npx install, no runtime, no MCP bindings. The AI is the…

★ 504Grade B
Gap Normalization✓SAFE

yogsoth-ai/de-anthropocentric-research-engine

A 267-skill research graph in pure markdown — 51 research operations built from 216 single-purpose steps, composed in any order with explicit backtracking. One npx install, no runtime, no MCP bindings. The AI is the…

★ 504Grade B
Gap Pairwise Judgment✓SAFE

yogsoth-ai/de-anthropocentric-research-engine

A 267-skill research graph in pure markdown — 51 research operations built from 216 single-purpose steps, composed in any order with explicit backtracking. One npx install, no runtime, no MCP bindings. The AI is the…

★ 504Grade B
Gap Prioritization✓SAFE

yogsoth-ai/de-anthropocentric-research-engine

A 267-skill research graph in pure markdown — 51 research operations built from 216 single-purpose steps, composed in any order with explicit backtracking. One npx install, no runtime, no MCP bindings. The AI is the…

★ 504Grade B
Gap Synthesis✓SAFE

yogsoth-ai/de-anthropocentric-research-engine

A 267-skill research graph in pure markdown — 51 research operations built from 216 single-purpose steps, composed in any order with explicit backtracking. One npx install, no runtime, no MCP bindings. The AI is the…

★ 504Grade B

Questions about Research Catalog

What does the Research Catalog skill do?
A 267-skill research graph in pure markdown — 51 research operations built from 216 single-purpose steps, composed in any order with explicit backtracking. One npx install, no runtime, no MCP bindings. The AI is the researcher; you set the direction.
Is Research Catalog safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can Research Catalog access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
How current is this page?
The grade is for one exact copy of the source (e94c5ee067cd), read on 2026-10-19. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.