Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →yogsoth-ai/de-anthropocentric-research-engine
Top-level orchestrator for the yogsoth-ai research ecosystem. Drives the full research lifecycle from direction crystallization through experiment design.
Passed 20 of 22 test cases with the skill, and 4 of 22 without it.
and 19 more in the full report
16 fixed4 passed either way2 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
This benchmark measured an earlier version of the skill.
One of 16 test cases the skill fixed. Same model, same prompt, with and without the skill.
Which SOP acts as the comprehensive capability menu and skill index for the 10 composable research packages?
| Checked | Without skill | With skill |
|---|---|---|
| Identifies research-catalog as the capability menu and skill index. | Missed | Passed |
The instructions your model reads when the skill loads.
DARE v4 is the product-level orchestrator for a research run. It turns a user request into a controlled sequence of North Star crystallization, executable specification, and phase-by-phase execution. It is not a scientific graph node and must not be counted among the 267 tactic/SOP nodes.
required: [user_request]
optional: [existing_context, north_star, research_brief, constraints]
constraints:
- a run may start with an existing North Star only when it is explicit and user-confirmed
- execution requires a current SpecView reconstructed from checkpoint events
DARE enforces this order and does not skip a phase:
research-catalog to expose the available tactics. You MUST load skill write-research-spec to turn the confirmed North Star, ResearchBrief, and user constraints into an executable Research Spec. Do not execute research while this phase is incomplete.execute-research-spec to execute its phases and checkpoints. Do not select a tactic directly from the flat skill directory.At every phase boundary, record the decision and its reason in the current Phase checkpoint event stream. A plan change is an appended decisions event, never an edit to an earlier checkpoint. The product shell may present progress, but scientific conclusions remain in tactic/SOP Delta fields.
produces:
- confirmed_north_star
- research_brief
- spec_view
- phase_execution_results
- final_research_summary
state_delta_fields:
- findings
- evidence_updates
- hypothesis_updates
- assumption_updates
test cases passed, vs 4 / 22 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.