antifragiletech/antifragile-claude-code/context-save
Claude Code and agent skill by antifragiletech/antifragile-claude-code
Trust tier 3
- community
Source check pending
Checking that the SKILL.md file is still in the repository.
Install
npx vskill@latest install antifragiletech/antifragile-claude-code/context-saveSKILL.md
/context-save — Context Efficiency Skill
Trigger
User says: "save context", "context-save", "be efficient", "don't waste context", or at the start of complex tasks
Purpose
Enforce strict context efficiency rules throughout the session.
Rules — ACTIVE FOR ENTIRE SESSION
File Reading
- Run
wc -lbefore reading ANY file - Files > 300 lines: NEVER read fully. Use grep → sed/head/tail
- After writing a file: NEVER re-read it
- After editing: NEVER re-read unless verifying a complex multi-step edit
Output Discipline
- NEVER echo back code you just wrote
- NEVER repeat file contents in your response
- Confirm writes in under 20 words: "Created [file] ([N] lines)."
- For edits: "Updated [file] line [N]: [what changed]."
- Skip preamble — no "I'll help you with that" or "Let me..."
Tool Call Efficiency
- Batch parallel reads when possible (multiple Read calls in one turn)
- Batch parallel greps when searching for related things
- Use Glob to find files, not Bash(find ...)
- Use Grep to search content, not Bash(grep ...)
- Prefer Edit over Write (don't rewrite entire files for small changes)
Session Management
- After every 5 tool calls, mentally check: "Am I being efficient?"
- If you've read the same file twice, you're doing it wrong
- If your response is > 200 words and doesn't contain code, it's too long
- Run /compact after creating 3+ files
Planning Discipline
- MAX 2 minutes in plan mode. If stuck, start coding immediately.
- No more than 1 round of file exploration before starting implementation
- If asked to fix a bug: grep → read section → fix. Max 3 tool calls.
- If asked for a feature: check 1-2 existing patterns → implement. Max 5 reads.
Anti-patterns to AVOID
- Reading a 500-line file to find one function → use grep first
- Re-reading a file after every edit → trust your edit
- "Let me explore the project structure..." → check memory first
- Running git status/diff when not asked → wastes a tool call
- Explaining your plan in detail before acting → just act
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingassets/skills/thinking/context-save/SKILL.md
Popularity
- 7-day trend score
- 5
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/antifragiletech/antifragile-claude-code/context-save)