sudocode-ai/everything-sudocode/strategic-compact

Claude Code and agent skill by sudocode-ai/everything-sudocode

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Suggests manual context compaction at logical intervals to preserve context through task phases rather than arbitrary auto-compaction.

Install

npx vskill@latest install sudocode-ai/everything-sudocode/strategic-compact

SKILL.md

Strategic Compact Skill

Suggests manual /compact at strategic points in your workflow rather than relying on arbitrary auto-compaction.

Why Strategic Compaction?

Auto-compaction triggers at arbitrary points:

  • Often mid-task, losing important context
  • No awareness of logical task boundaries
  • Can interrupt complex multi-step operations

Strategic compaction at logical boundaries:

  • After exploration, before execution - Compact research context, keep implementation plan
  • After completing a milestone - Fresh start for next phase
  • Before major context shifts - Clear exploration context before different task

How It Works

The suggest-compact.sh script runs on PreToolUse (Edit/Write) and:

  1. Tracks tool calls - Counts tool invocations in session
  2. Threshold detection - Suggests at configurable threshold (default: 50 calls)
  3. Periodic reminders - Reminds every 25 calls after threshold

Hook Setup

Add to your ~/.claude/settings.json:

json
{
"hooks": {
"PreToolUse": [{
"matcher": "tool == \"Edit\" || tool == \"Write\"",
"hooks": [{
"type": "command",
"command": "~/.claude/skills/strategic-compact/suggest-compact.sh"
}]
}]
}
}

Configuration

Environment variables:

  • COMPACT_THRESHOLD - Tool calls before first suggestion (default: 50)

Best Practices

  1. Compact after planning - Once plan is finalized, compact to start fresh
  2. Compact after debugging - Clear error-resolution context before continuing
  3. Don't compact mid-implementation - Preserve context for related changes
  4. Read the suggestion - The hook tells you when, you decide if

Sudocode Integration

Compaction Boundaries in Sudocode Workflow

The sudocode workflow has natural compaction boundaries:

Phase 1: Capture (Spec) ← Compact AFTER creating spec
Phase 2: Plan (Issue) ← Compact AFTER planning issues
Phase 3: Execute (Tasks) ← DON'T compact mid-implementation
Phase 4: Close (Feedback) ← Compact AFTER closing issue
Recommended Compaction Points

Good times to compact:

  1. After spec creation
  • Spec is persisted in .sudocode/specs/
  • Context can be reloaded with show_spec
  • Compact to start fresh for planning
  1. After issue planning
  • Issues are persisted with relationships
  • ready will show what to work on
  • Compact before implementation
  1. After closing an issue
  • Work is complete and documented
  • Feedback is attached to spec
  • Fresh context for next issue

Bad times to compact:

  1. Mid-implementation
  • Lose context about current approach
  • May duplicate work or miss edge cases
  1. Before providing feedback
  • Need context to write good feedback
  • Wait until feedback is submitted
  1. While debugging
  • Investigation context is crucial
  • Compact after resolution
Sudocode-Aware Suggestion

Update the compaction suggestion to check sudocode state:

bash
# In suggest-compact.sh, add check:
CURRENT_ISSUE=$(sudocode issue list --status=in_progress --format=count 2>/dev/null)
if [ "$CURRENT_ISSUE" -gt 0 ]; then
echo "⚠️ Issue in progress - consider completing before compact"
else
echo "✓ No issue in progress - safe to compact"
fi
Context Preservation Across Compaction

Since sudocode persists context, after compaction you can restore:

1. /compact
2. ready # See available work
3. show_issue <id> # Reload issue context
4. show_spec <linked-spec> # Reload requirements
5. Continue implementation

Key context survives in:

  • .sudocode/specs/ - All spec content
  • .sudocode/issues/ - Issue state and relationships
  • Feedback - Attached to specs
  • Git history - Code changes

Security

  • Security scanPassed, 95 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

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
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/sudocode-ai/everything-sudocode/strategic-compact/badge)](https://verified-skill.com/skills/sudocode-ai/everything-sudocode/strategic-compact)

Something wrong with this skill or its page? Tell us.