llm context skills

18 tagged llm context, measured the same way as everything else here.

Browse within: ai-instructions 12context 12context-engineering 12instructions 12

codestory-grounding

01

TheGreenCedar/CodeStory

Skill Claude CodeCodex

Use when an agent should ground a local repository with CodeStory before making source claims, planning edits, choosing tests, reviewing changes, or using broad retrieval evidence through the CodeStory plugin MCP.

37 2d ago A 44 tokens original Apache-2.0

adopt-memory-bank

02

mehmetdemirci/comb-ai

Skill Claude CodeCodex

Add a COMB memory bank to an existing (brownfield) project by analyzing current codebase, dependencies, and architecture. Use when the project already has code but no memory bank.

13 6mo ago A 42 tokens original MIT

create-memory-bank

03

mehmetdemirci/comb-ai

Skill Claude CodeCodex

Initialize a COMB memory bank for a new (greenfield) project. Creates the directory structure, core files, and INDEX.md from templates. For existing projects, use /adopt-memory-bank instead.

13 6mo ago A 45 tokens original MIT

update-memory-bank

04

mehmetdemirci/comb-ai

Skill Claude CodeCodex

Update memory bank after task completion. Detects scope automatically and updates files in priority order. Can be auto-invoked when task completion is detected in conversation.

13 6mo ago A 36 tokens original MIT

create-checkpoint

05

MSiccDev/ai-context-kit

Skill Claude CodeCodex

Capture the current session state as a checkpoint artifact per spec section 4.4. Use when the user signals session end or explicitly requests a checkpoint. Do not use mid-session or without user approval.

12 4d ago A 44 tokens original MIT

obsidian-kb

08

chiKeka/obsidian-kb

Skill Claude CodeCodex

Use this skill when the user wants to build a knowledge base for their project, create an Obsidian wiki from project data, analyze project structure for knowledge management, or mentions "obsidian-kb", "knowledge base", or "compile wiki".

4 5mo ago A 55 tokens original MIT

map-token-context

09

hakiyaka/capsule

Skill Claude CodeCodex

Compile branching decisions outside model generation, route virtualized specialist skills, and minimize model input/output across tools, files, history, images, subagents, search, and recall. Use for token-heavy or reasoning-heavy work or reversible skill-vault management.

1 22d ago A 54 tokens original MIT