Back to list
imrellx

deciduous

by imrellx

A dark Obsidian theme combining Jellybeans colors with Flexoki backgrounds

1🍴 0📅 Jan 2, 2026

SKILL.md


name: deciduous description: Plan, implement, track, and reflect on your work goals and decisions.

Planning & Decision Graph Logging

Track every goal, decision, and outcome in the decision graph. This creates persistent memory that survives context loss.

  • ALWAYS LOG BEFORE YOU CODE, NOT AFTER.
  • Log at the granularity of TODOs or task items.
  • When drafting a plan create the GOAL node.
  • User Decisions should be tracked

When to Log (Automatic Triggers)

SituationNode TypeExample
In plan modegoal"Add user authentication"
TODO / Task Itemaction"Implementing JWT auth middleware"
User requests new featuregoal"Add user authentication"
Choosing between approachesdecision"Choose between JWT vs sessions"
Considering an optionoption"Use JWT with refresh tokens"
About to write/edit codeaction"Implementing JWT auth middleware"
Work completed or failedoutcome"JWT auth working" or "JWT approach failed"
Important observationobservation"Existing code uses cookie-based sessions"

Commands

# Create nodes (always include confidence -c)
deciduous add goal "Title" -c 90 -p "User's exact request"
deciduous add decision "Title" -c 75
deciduous add action "Title" -c 85
deciduous add outcome "Title" -c 90
deciduous add observation "Title" -c 80

# CRITICAL: Link nodes immediately after creation
deciduous link <parent_id> <child_id> -r "Reason for connection"

# After git commits, link to the graph
deciduous add action "Committed feature X" -c 90 --commit HEAD

# View the graph
deciduous nodes
deciduous edges

Rules

  1. Log BEFORE acting - Create the action node before writing code
  2. Link IMMEDIATELY - Every node except root goals must have a parent
  3. Capture verbatim prompts - Use -p with the user's exact words for goals
  4. Include confidence - Always use -c flag (0-100)
  5. Log outcomes - Both successes AND failures get logged

Confidence Guidelines

  • 90-100: Certain, verified, tested
  • 75-89: High confidence, likely correct
  • 50-74: Moderate confidence, some uncertainty
  • Below 50: Experimental, speculative

The Memory Loop

User Request → Log goal with -p
    ↓
Choose Approach → Log decision + options
    ↓
Start Coding → Log action FIRST
    ↓
Complete Work → Log outcome, link to parent
    ↓
Git Commit → Log with --commit HEAD

Remember: The decision graph is your persistent memory. Log as you work, not after.

Score

Total Score

65/100

Based on repository quality metrics

SKILL.md

SKILL.mdファイルが含まれている

+20
LICENSE

ライセンスが設定されている

+10
説明文

100文字以上の説明がある

0/10
人気

GitHub Stars 100以上

0/15
最近の活動

3ヶ月以内に更新

+5
フォーク

10回以上フォークされている

0/5
Issue管理

オープンIssueが50未満

+5
言語

プログラミング言語が設定されている

+5
タグ

1つ以上のタグが設定されている

+5

Reviews

💬

Reviews coming soon