Other

conflict-of-interest-checker

Check for co-authorship and institutional conflicts between authors and suggested reviewers to support peer review integrity. Coauthorship and institutional conflict detection supported.

88100Total Score
Core Capability
87 / 100
Functional Suitability
11 / 12
Reliability
11 / 12
Performance & Context
7 / 8
Agent Usability
15 / 16
Human Usability
7 / 8
Security
10 / 12
Maintainability
12 / 12
Agent-Specific
14 / 20
Medical Task
12 / 12 Passed
89Check coauthorship conflicts for Smith/Jones/Lee vs Brown/Davis/Wilson
4/4
87Check conflicts using a custom CSV publication file
4/4
90Check institutional conflict between reviewer and author at same institution
4/4

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

Core Capability87 / 100 — 8 Categories

Functional Suitability
TypeError (list & list) fixed in v4: set() conversion applied in check_coauthorship_conflict(). NameError in check_institutional_conflict() fixed: 'reviewer' replaced with reviewer_name parameter. Collaboration window stub still documented as known limitation.
11 / 12
92%
Reliability
Both runtime bugs fixed. CSV error handling now uses try/except with clear FileNotFoundError and generic Exception messages. Script runs cleanly on canonical input without fallback required.
11 / 12
92%
Performance & Context
SKILL.md is 145 lines — concise. Script is minimal at 172 lines. Very low token cost.
7 / 8
88%
Agent Usability
Script now executes correctly for coauthorship detection without requiring Claude Direct fallback. Institutional conflict path still documented as Claude Direct but script method is now also fixed. Error handling rules updated to reflect fixed state.
15 / 16
94%
Human Usability
Description is discoverable for journal submission and editorial use cases. Scope boundaries are well-defined.
7 / 8
88%
Security
No credential handling. Demo data hardcoded in script is acceptable for testing. No injection vectors. CSV parsing uses DictReader safely.
10 / 12
83%
Maintainability
Clean single-file script. Both bugs fixed. Known limitations documented. CSV error handling added. Collaboration window stub documented.
12 / 12
100%
Agent-Specific
Trigger description is precise. Escape hatches present. Institutional conflict path documented. Script now executes correctly for primary use case. Collaboration window stub still unimplemented — documented gap.
14 / 20
70%
Core Capability Total87 / 100

Medical TaskExecution Average: 88.7 / 100 — Assertions: 12/12 Passed

89
Canonical
Check coauthorship conflicts for Smith/Jones/Lee vs Brown/Davis/Wilson
4/4 ✓
87
Variant A
Check conflicts using a custom CSV publication file
4/4 ✓
90
Edge
Check institutional conflict between reviewer and author at same institution
4/4 ✓
89
Canonical✅ Pass
Check coauthorship conflicts for Smith/Jones/Lee vs Brown/Davis/Wilson

Script executes without error. TypeError fix confirmed: set() conversion in check_coauthorship_conflict() resolves list & list intersection. Correctly identifies 3 conflicts: Brown-Smith (paper1), Wilson-Smith (paper2), Wilson-Jones (paper2).

Basic 36/40|Specialized 53/60|Total 89/100
✅A1Script executes without runtime error on canonical input
✅A2Output identifies coauthorship conflicts correctly
✅A3Output includes Accept/Recuse recommendation for each reviewer
✅A4Output stays within conflict-of-interest checking scope
Pass rate: 4 / 4
87
Variant A✅ Pass
Check conflicts using a custom CSV publication file

CSV-based conflict detection works correctly. CSV error handling confirmed: FileNotFoundError produces clear error message and exits cleanly. Valid CSV correctly loads and detects conflicts.

Basic 35/40|Specialized 52/60|Total 87/100
✅A1Output loads CSV correctly and uses it instead of demo data
✅A2Output correctly identifies conflicts from CSV records
✅A3CSV error handling produces clear message on missing file
✅A4Output does not exceed stated scope
Pass rate: 4 / 4
90
Edge✅ Pass
Check institutional conflict between reviewer and author at same institution

NameError fix confirmed: check_institutional_conflict() now uses reviewer_name parameter correctly. Script method works. Claude Direct path also documented as alternative. Both paths produce correct institutional conflict detection.

Basic 36/40|Specialized 54/60|Total 90/100
✅A1check_institutional_conflict() executes without NameError
✅A2Institutional conflict is detected when reviewer and author share the same institution
✅A3Output reports institutional conflict type separately from coauthorship
✅A4Output does not exceed stated scope
Pass rate: 4 / 4
Medical Task Total88.7 / 100

Key Strengths

  • TypeError (list & list) fixed: set() conversion in check_coauthorship_conflict() enables correct script execution on canonical input
  • NameError in check_institutional_conflict() fixed: reviewer_name parameter now used correctly, enabling script-level institutional conflict detection
  • CSV error handling added: FileNotFoundError and generic exceptions caught with clear error messages
  • Collaboration window stub clearly documented as known limitation with no silent failure