AgentSkillsCN

Aws Database Rules

AWS DynamoDB PlayerValue 表的交互规则与数据结构规范。

SKILL.md
--- frontmatter
description: Rules and schema for interacting with the AWS DynamoDB PlayerValue table.

Skill: AWS Database Rules

Use this skill to understand the data model and access patterns for the application's main DynamoDB table.

Table Information

  • •Table Name Pattern: PlayerValue-* (Retrieve the full name from amplify_outputs.json or environment variables).
  • •Primary Key (Partition Key): sleeper_id (String).
  • •Sort Key: None.

Schema Fields (PlayerValue)

The PlayerValue table stores enriched player data, combining static analysis with dynamic market data.

FieldTypeDescription
sleeper_idStringPK. Unique ID from Sleeper API. Use this for lookups.
player_name_originalStringPlayer's original name from source data.
full_nameStringPlayer's full name (added by playerService as alias for player_name_original).
player_cleansed_nameStringCleansed name for matching (lowercase, no punctuation).
positionStringQB, RB, WR, TE, etc.
teamStringNFL Team Abbreviation (e.g., MIN, KC).
fantasy_calc_valueNumberTrade value from FantasyCalc API (Market Value).
fc_rankNumberMarket Rank from FantasyCalc.
overall_rankNumberUser's SF Rank. From Superflex spreadsheet Overall column.
one_qb_rankNumberUser's 1QB Rank. From 1QB spreadsheet Overall column.
positional_rankNumberUser's Positional Rank (SF). From Positional Rank column.
one_qb_pos_rankNumberUser's Positional Rank (1QB). From Positional Rank column.
tierNumberUser's Tier (SF). From Tier column.
one_qb_tierNumberUser's Tier (1QB). From Tier column.
trend_30_dayNumber30-day value trend from FantasyCalc.
redraft_valueNumberRedraft value from FantasyCalc.

Important Field Mappings

  • •playerService.js: Adds full_name as alias for player_name_original when loading from DynamoDB
  • •Matching: Use player_cleansed_name for name-based lookups (removes punctuation, lowercase)
  • •Rankings: overall_rank and one_qb_rank come from uploaded Excel files, NOT FantasyCalc

Access Patterns

  1. •Batch Get: Use BatchGetItem with a list of sleeper_ids to fetch data for a roster.
  2. •Name Lookup: Avoid explicit scans. Use cleanseName to normalize names if ID lookup fails.