Babysitter context-preservation

State capture and restore across context window compactions. Monitors usage thresholds and serializes quality, task, and spec state for seamless continuation.

install

source · Clone the upstream repo

git clone https://github.com/a5c-ai/babysitter

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/a5c-ai/babysitter "$T" && mkdir -p ~/.claude/skills && cp -r "$T/library/methodologies/pilot-shell/skills/context-preservation" ~/.claude/skills/a5c-ai-babysitter-context-preservation && rm -rf "$T"

manifest: library/methodologies/pilot-shell/skills/context-preservation/SKILL.md

context-preservation

You are context-preservation -- the state preservation skill for Pilot Shell.

Overview

This skill manages context window state across compactions, ensuring seamless continuation of work when context is refreshed. It implements the PreCompact and post_compact_restore patterns from Pilot Shell.

Capabilities

1. State Capture (PreCompact)

Serialize current spec state (tasks, statuses, acceptance criteria)
Capture quality pipeline state (last lint/format/typecheck results)
Save TDD progress (current phase, iteration count, scores)
Store context monitor metrics
Write to
```
.pilot-shell/state.json
```

2. State Restore (SessionStart / post_compact_restore)

Read
```
.pilot-shell/state.json
```
on session start
Restore spec task tracking state
Restore quality baseline
Resume TDD from last known position
Log restoration summary

3. Threshold Monitoring

Track context usage percentage (default threshold: 70%)
Trigger preservation when threshold approached
Calculate optimal preservation timing

State Schema

{
  "version": "1.0.0",
  "timestamp": "2026-03-02T12:00:00Z",
  "spec": {
    "title": "...",
    "taskStatuses": [{ "id": "...", "status": "COMPLETE" }],
    "currentPhase": "implement"
  },
  "quality": {
    "lastScore": 87,
    "lint": { "passed": true },
    "format": { "passed": true },
    "typecheck": { "passed": true }
  },
  "tdd": {
    "iteration": 2,
    "score": 92,
    "compliant": true
  },
  "context": {
    "usagePercent": 72,
    "preservedAt": "2026-03-02T12:00:00Z"
  }
}