Recommendation Policy Schema

Reference for recommendation.toml used by:

• honestroles recommend build-index --policy recommendation.toml
• honestroles recommend match --policy recommendation.toml
• honestroles recommend evaluate --policy recommendation.toml

If omitted, HonestRoles uses the built-in default scoring policy.

Schema

[weights]
skills = 0.35
title = 0.20
location_work_mode = 0.15
seniority = 0.10
recency = 0.10
compensation = 0.10

reason_limit = 3

Fields

Field	Type	Default	Notes
weights.skills	float >= 0	0.35	Skill-overlap signal weight
weights.title	float >= 0	0.20	Title-similarity signal weight
weights.location_work_mode	float >= 0	0.15	Location/work-mode fit weight
weights.seniority	float >= 0	0.10	Seniority-fit weight
weights.recency	float >= 0	0.10	Posting recency weight
weights.compensation	float >= 0	0.10	Compensation-fit weight
reason_limit	integer >= 1	3	Number of top match_reasons returned

Weights are normalized to sum to 1 at runtime.