con un clic
archive-prd
Archives completed stories from prd.json to reduce token usage.
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Menú
Archives completed stories from prd.json to reduce token usage.
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Basado en la clasificación ocupacional SOC
| name | archive-prd |
| description | Archives completed stories from prd.json to reduce token usage. |
| triggers | ["archive","compact prd","prd too large"] |
| allowed-tools | Read, Write, Edit, Bash |
| model | haiku |
| user-invocable | true |
| argument-hint | [status|S-ID|unarchive S-ID] |
Archive completed stories to keep prd.json fast and small. Keep only last 3 sprints active.
1. Read prd.json
2. Separate stories:
- ACTIVE: passes=false OR passes=null OR type="qa"
- COMPLETED: passes=true AND type!="qa"
3. Create archive file: .claude/archives/prd-archive-YYYY-MM.json
4. Update main prd.json with summary
5. Report: "Archived X stories, Y remain active"
{
"project": "Project Name",
"version": "1.3.2",
"lastUpdated": "2026-01-22",
"roadmapPhase": "Current Phase",
"archived": {
"totalCompleted": 41,
"lastArchived": "2026-01-22",
"files": [".claude/archives/prd-archive-2026-01.json"],
"summary": {
"S01-S10": "Core foundation - registry, funnels, OAuth, caching",
"S11-S20": "Navigation, QA, dashboard, exports, favorites",
"S21-S30": "Time granularity, GA4 schema, accessibility, mobile",
"S31-S41": "Token refresh, metrics, documentation, scope rules"
}
},
"stories": [
// Only active/pending stories here
]
}
{
"archivedAt": "2026-01-22T10:00:00Z",
"project": "Project Name",
"version": "1.3.2",
"stories": [
// Full story objects for reference
]
}
| Condition | Action |
|---|---|
| 4+ total sprints | Auto-suggest archive |
| prd.json > 500 lines | Suggest archive |
| prd.json > 50KB | Force archive |
| User says "archive" | Manual archive |
| All stories complete | Archive and start fresh |
Keep only the last 3 sprints in prd.json. Archive everything older.
1. BACKUP
mkdir -p .claude/archives
cp prd.json .claude/archives/prd-backup-$(date +%Y%m%d).json
2. EXTRACT COMPLETED
Filter: passes=true AND type!="qa"
3. CREATE ARCHIVE
Write to: .claude/archives/prd-archive-YYYY-MM.json
4. GENERATE SUMMARY
Group stories by ID range (10 per group)
Write 1-line summary per group
5. UPDATE MAIN PRD
Remove archived stories
Add "archived" section with summary
Keep all QA stories (even passed ones for re-testing)
6. VALIDATE
Ensure main prd.json < 1500 lines
Ensure all story IDs accounted for
If you need details on an archived story:
User: "What was S15 about?"
Claude:
1. Check archived.summary for S15 range
2. Read .claude/archives/prd-archive-2026-01.json if needed
3. Report story details
| Say | Action |
|---|---|
archive | Archive completed stories |
archive status | Show archive stats |
archive S15 | Show archived story S15 |
unarchive S15 | Restore story to active |
| State | Estimated Tokens |
|---|---|
| Full prd.json (70+ stories, 10 sprints) | ~25,000+ |
| After archive (current sprint + 2) | ~3,000-5,000 |
| With summary | +500 |
| Total Savings | ~80% |
Real example: a long-running project can grow to 800+ rows / 10 sprints / ~20K tokens. After archiving to keep only the current sprint + 2 prior, prd.json drops to ~150 rows / ~3K tokens.
Show token / tool usage stats from the local telemetry log. Use when you want to know "which tools am I burning context on", "which skills are expensive", or "was yesterday's session mostly Read/Grep or actually productive".
Parallel quality audit with 7 specialized agents (Opus). Finds bugs, violations, and quality issues. Use audit for fixes, brainstorm for features.
Manage environment variables with Doppler — auto-install CLI, login, link projects, wrap commands with `doppler run`. Replaces scattered .env files with a hub/spoke architecture.
Scaffolds new projects or onboards existing ones. Detects stack, creates monorepo/single-app, configures strict tooling. Use for greenfield or first-time setup.
Autonomous task execution with testing and security. Works through all tasks without stopping.
Removes temporary screenshots, old backups, stale handoffs, and auto-active flags. Use when project has accumulated temp files.