Evidence Insight

diffdock-molecular-docking

85100Total Score
Core Capability
85 / 100
Functional Suitability
11 / 12
Reliability
9 / 12
Performance & Context
8 / 8
Agent Usability
14 / 16
Human Usability
8 / 8
Security
9 / 12
Maintainability
9 / 12
Agent-Specific
17 / 20
Medical Task
15 / 20 Passed
86Blind docking when you have a protein structure (PDB) and a ligand (SMILES) but no known binding site
3/4
85Pose prediction to generate multiple plausible 3D binding conformations and rank them
3/4
85Diffusion generative sampling to produce diverse ligand binding poses
3/4
85Confidence model scoring to rank predicted poses
3/4
85End-to-end case for Diffusion generative sampling to produce diverse ligand binding poses
3/4

Veto GatesRequired pass for any deployment consideration

Skill Veto✓ All 4 gates passed
✓
Operational Stability
System remains stable across varied inputs and edge cases
PASS
✓
Structural Consistency
Output structure conforms to expected skill contract format
PASS
✓
Result Determinism
Equivalent inputs produce semantically equivalent outputs
PASS
✓
System Security
No prompt injection, data leakage, or unsafe tool use detected
PASS
Research Veto✅ PASS — Applicable
DimensionResultDetail
Scientific IntegrityPASSThe legacy review kept diffdock-molecular-docking tied to supported scientific content rather than invented findings.
Practice BoundariesPASSThe package stayed in retrieval, extraction, or evidence-organization scope rather than drifting into unsupported interpretation.
Methodological GroundPASSThe archived evaluation treated the workflow as method-linked rather than ad hoc.
Code UsabilityPASSCode usability passed because the package still exposed a reviewable execution surface for its documented workflow.

Core Capability85 / 100 — 8 Categories

Functional Suitability
The archived deduction in functional suitability traces back to: Improve stress-case output rigor. Stress and boundary scenarios show weaker consistency
11 / 12
92%
Reliability
Related legacy finding for diffdock-molecular-docking: Improve stress-case output rigor. Stress and boundary scenarios show weaker consistency
9 / 12
75%
Performance & Context
The legacy audit gave full marks to performance context for this package.
8 / 8
100%
Agent Usability
The legacy audit deducted points for diffdock-molecular-docking in agent usability.
14 / 16
88%
Human Usability
The legacy audit gave full marks to human usability for this package.
8 / 8
100%
Security
The legacy audit deducted points for diffdock-molecular-docking in security.
9 / 12
75%
Maintainability
The archived evaluation left some headroom for diffdock-molecular-docking under maintainability.
9 / 12
75%
Agent-Specific
Related legacy finding for diffdock-molecular-docking: Stabilize executable path and fallback behavior. Some inputs only reached PARTIAL due to execution gaps or weak boundary handling
17 / 20
85%
Core Capability Total85 / 100

Medical TaskExecution Average: 85.2 / 100 — Assertions: 15/20 Passed

86
Canonical
Blind docking when you have a protein structure (PDB) and a ligand (SMILES) but no known binding site
3/4 ✓
85
Variant A
Pose prediction to generate multiple plausible 3D binding conformations and rank them
3/4 ✓
85
Edge
Diffusion generative sampling to produce diverse ligand binding poses
3/4 ✓
85
Variant B
Confidence model scoring to rank predicted poses
3/4 ✓
85
Stress
End-to-end case for Diffusion generative sampling to produce diverse ligand binding poses
3/4 ✓
86
Canonical✅ Pass
Blind docking when you have a protein structure (PDB) and a ligand (SMILES) but no known binding site

The archived evidence for Blind docking when you have a protein structure (PDB) and a ligand... shows a real execution snag, though not one that erased the workflow contract.

Basic 33/40|Specialized 53/60|Total 86/100
✅A1The diffdock-molecular-docking output structure matches the documented deliverable
❌A2The script execution path completed successfully for the documented case
✅A3The output stays fully within the documented skill boundary
✅A4The response quality is acceptable for the documented path
Pass rate: 3 / 4
85
Variant A✅ Pass
Pose prediction to generate multiple plausible 3D binding conformations and rank them

Pose prediction to generate multiple plausible 3D binding... kept a coherent task boundary even though the preserved run encountered a recoverable error.

Basic 31/40|Specialized 54/60|Total 85/100
✅A1The diffdock-molecular-docking output structure matches the documented deliverable
❌A2The script execution path completed successfully for the documented case
✅A3The output stays fully within the documented skill boundary
✅A4The response quality is acceptable for the documented path
Pass rate: 3 / 4
85
Edge✅ Pass
Diffusion generative sampling to produce diverse ligand binding poses

The archived evidence for Diffusion generative sampling to produce diverse ligand binding poses shows a real execution snag, though not one that erased the workflow contract.

Basic 30/40|Specialized 55/60|Total 85/100
✅A1The diffdock-molecular-docking output structure matches the documented deliverable
❌A2The script execution path completed successfully for the documented case
✅A3The output stays fully within the documented skill boundary
✅A4The response quality is acceptable for the documented path
Pass rate: 3 / 4
85
Variant B✅ Pass
Confidence model scoring to rank predicted poses

Confidence model scoring to rank predicted poses kept a coherent task boundary even though the preserved run encountered a recoverable error.

Basic 29/40|Specialized 56/60|Total 85/100
✅A1The diffdock-molecular-docking output structure matches the documented deliverable
❌A2The script execution path completed successfully for the documented case
✅A3The output stays fully within the documented skill boundary
✅A4The response quality is acceptable for the documented path
Pass rate: 3 / 4
85
Stress✅ Pass
End-to-end case for Diffusion generative sampling to produce diverse ligand binding poses

End-to-end case for Diffusion generative sampling to produce... kept a coherent task boundary even though the preserved run encountered a recoverable error.

Basic 26/40|Specialized 59/60|Total 85/100
✅A1The diffdock-molecular-docking output structure matches the documented deliverable
❌A2The script execution path completed successfully for the documented case
✅A3The output stays fully within the documented skill boundary
✅A4The response quality is acceptable for the documented path
Pass rate: 3 / 4
Medical Task Total85.2 / 100

Key Strengths

  • Primary routing is Evidence Insight with execution mode B
  • Static quality score is 85/100 and dynamic average is 73.6/100
  • Assertions and command execution outcomes are recorded per input for human review
  • Execution verification summary: Script verification 1/2; adjustment=3. inference_runner.py: OK; setup_check.py: rc=1