specialization-update

name	specialization-update
description	Process for updating operator specialization (专精) training efficiency data. Used when new operators are released or existing training room buffs change. Updates maximum_efficiency.json and optimal_specialization.json.

Overview

This skill covers the end-to-end process of updating 专精 (specialization/mastery training) related JSON data files when new operators are released with training room buffs, or when existing operators' base skills change.

Key Files

File	Purpose
src/static/json/build/building_table.json	Raw base skill data; source of truth for all training room buffs
src/static/json/build/maximum_efficiency.json	Complete ranking of every operator's training room efficiency
src/static/json/build/optimal_specialization.json	Best assistant per profession/branch/mastery-level combination
src/static/json/operator/character_table_simple.json	Operator basic info (name, profession, subProfessionId)

Process

Step 1: Identify the operator's training buffs

Search building_table.json for the operator by name or charId. Filter to entries where roomType is "training". Note:

• phase 0 = base skill (usually lower efficiency, unlocked at E0)
• phase 2 = upgraded skill (higher efficiency, unlocked at E2)
• The phase: 2 entry gives the max efficiency
• Extract the percentage from the description field

Step 2: Determine max efficiency value

Parse the training speed percentage from the description and convert to decimal:

• +30% → 0.3
• +45% → 0.45
• +70% → 0.7
• Conditional buffs should use the maximum achievable value

Step 3: Determine profession coverage

From the description, identify which professions the buff applies to:

• Single profession: e.g. "profession": "近卫"
• Multiple professions: e.g. "profession": "术师/近卫", "conditions": { "target_profession": ["术师", "近卫"] }
• All professions: "profession": "通用" (no target_profession needed)
• Branch-specific: e.g. "branch": "领主" (noted in conditions)

Step 4: Add to maximum_efficiency.json

Add a new entry following this template:

Flat buff (no mastery level condition):

{
  "name": "<干员名>",
  "profession": "<职业1>/<职业2>",
  "max_efficiency": <小数>,
  "conditions": {
    "target_profession": ["<职业1>", "<职业2>"],
    "skill_name": "<技能名>"
  },
  "skill_name": "<完整技能描述>",
  "charId": "<char_xxxx_xxxx>"
}

Conditional buff (specific mastery level):

{
  "name": "<干员名>",
  "profession": "<职业>",
  "max_efficiency": <总效率>,
  "conditions": {
    "target_profession": "<职业>",
    "skill_name": "<技能名>",
    "mastery_level": <1|2|3>,
    "efficiency_base": <基础效率>,
    "efficiency_bonus": <额外效率>
  },
  "skill_name": "<完整技能描述>",
  "charId": "<char_xxxx_xxxx>"
}

Branch-specific buff:

{
  "name": "<干员名>",
  "profession": "<职业>",
  "max_efficiency": <总效率>,
  "conditions": {
    "target_profession": "<职业>",
    "skill_name": "<技能名>",
    "branch": "<分支名>",
    "efficiency_base": <基础效率>,
    "efficiency_bonus": <额外效率>
  },
  "skill_name": "<完整技能描述>",
  "charId": "<char_xxxx_xxxx>"
}

Insert the new entry at the top of the array.

Step 5: Check if optimal_specialization.json needs updating

Compare the new operator's efficiency against the best_assistant in each relevant slot in optimal_specialization.json. The file structure:

professions → <profession> → general → specialization_<N> → best_assistant
professions → <profession> → branches → <branch_name> → best_assistant

For each profession/branch that the new operator covers, check:

1. If the new efficiency is higher than the current best_assistant.efficiency, replace it
2. Consider mastery level conditions (specialization_1/2/3) separately
3. If the buff only applies at a specific mastery level, only update that level's entry
4. If the buff is unconditional, compare against all levels

Do NOT update if the new efficiency is lower or equal to the existing best.

Step 6: Update UI component branch list

When a new branch is added to optimal_specialization.json (e.g. 游击手 under 辅助), the corresponding UI component must also be updated.

Edit src/components/specialization/OptimalSpecializationPaths.vue and add the new branch to the branchesByProfession object for the matching profession:

const branchesByProfession = {
  "近卫": ["领主", "斗士"],
  "狙击": ["攻城手", "速射手"],
  "术师": [],
  "医疗": ["链愈师", "行医"],
  "重装": ["驭法铁卫"],
  "辅助": ["游击手"],  // ← add new branch here
  "特种": [],
  "先锋": []
};

If the profession already has other branches, simply append the new branch to the existing array. If the profession has no existing branches, create a new array with the branch name.

Do NOT update if the branch already exists in the branchesByProfession object.

Step 7: Commit

Use the commit format:

update(specialization): 职业专精最优路线和专精效率排行榜更新

Examples

Example 1: Operator with conditional specialization_3 buff (望, char_2027_wang)

Buff: "training speed +70% when training to mastery 3" (any profession)

Added to maximum_efficiency.json as first entry (new operator). Updated optimal_specialization.json:

• 医疗 → general → specialization_3: replaced 阿 (0.6) with 望 (0.7)
• 重装 → general → specialization_3: replaced 星熊 (0.6) with 望 (0.7)

Example 2: Operator with flat buff that doesn't beat any best (维伊, char_4226_veen)

Buff: "术师与近卫 training speed +45%" (flat, E2)

Only added to maximum_efficiency.json. Not added to optimal_specialization.json because existing best assistants for 术師/近卫 all have 0.75+ efficiency.

Example 3: Operator with branch-specific buff

Buff: "近卫 training speed +30%, +45% extra if branch is 领主" (仇白)

• profession: "近卫", max_efficiency: 0.75
• Only updates the branches.领主 entry, not general

Common Patterns

Buff Type	max_efficiency	Profession Label	conditions Pattern
Single profession flat +β (50%)	0.5	Single name	target_profession only
Single profession +α (30%)	0.3	Single name	target_profession only
Single profession +α/β + mastery/branch bonus	0.75	Single name	target_profession + branch or mastery_level + efficiency_base + efficiency_bonus
Single profession +α + mastery bonus (65%)	0.95	Single name	target_profession + mastery_level + efficiency_base + efficiency_bonus
Dual profession flat	varies	"X/Y"	target_profession array
All professions flat	0.25	"通用"	skill_name only
All professions conditional	varies	"通用"	skill-specific conditions

Edge Cases

1. Half-time operators (艾丽妮/逻各斯): These reduce training time by 50% via a separate mechanic. They are listed in optimal_specialization.json under half_time_operators, not in the professions tree.

2. Stacking buffs (耀骑士临光, 导火索 etc.): Operators whose buff scales with number of specific faction operators in base. Use the maximum achievable value as max_efficiency.

3. Special mechanics (余): Buffs that scale with special resources (e.g. 人间烟火). Note the special condition in conditions.special.

4. Operators without training buffs: Skip these entirely — they should NOT be added to either file.