# Sf Rebalance

> Refresh the deployed bot roster — pause decaying bots AND propose fresh top picks to replace them on the same exchanges with the same sizing. Use when the user asks "rebalance my bots / swap losers for winners / refresh the roster / clean up and replace / optimize my deployed strategies". Writes state (pauses) only with --apply; never auto-deploys.

- Skill: `daviddme/sf-rebalance` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add daviddme/sf-rebalance`
- Raw SKILL.md: https://api.skillmd.com/api/skills/daviddme/sf-rebalance/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: daviddme (https://skillmd.com/u/daviddme)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/daviddme/sf-rebalance

---


# sf-rebalance

The two-step roster refresh. Combines `sf-prune` (find decaying bots) and
`sf-best-pick` (find replacement strategies) into a single proposal you
can scan on stream and execute in one shot.

## What it does

1. Audit every ACTIVE bot's underlying strategy for edge decay.
2. For each `decaying` or `dead` bot, propose a replacement from the
   current top-N composite-score picks (filtered to bots NOT already in
   your roster, to avoid duplicates).
3. Optionally `--apply` the pauses.
4. NEVER auto-creates new bots — that's `sf-deploy`'s job. Replacements
   are surfaced as ready-to-run `sf-deploy` commands.

## Why not auto-deploy?

Pausing a bot is reversible. Creating one isn't quite — you'd have to
pause-then-delete (which isn't an endpoint). So this skill stays on the
careful side: prune now, propose replacement, let the operator pull the
trigger on the new deploy.

## When to use

- "Rebalance my bots"
- "Swap the losers for winners"
- "Refresh the roster — kill what's dead, suggest replacements"

## How to run

```
# Pure proposal mode — no changes
python sf-rebalance/rebalance.py [--top-picks N] [--json]

# Pause the dying ones; still only PROPOSES replacements
python sf-rebalance/rebalance.py --apply [--top-picks N]
```

## Output

Two sections:

```
TO PAUSE
  [decaying] BotName  strategy=X  flags=...
  [dead]     BotName  strategy=Y  flags=...

REPLACEMENT CANDIDATES (run sf-deploy to actually create)
  --strategy slug-a --exchange EX_ID --position percentage --amount 3 --leverage 10
  --strategy slug-b --exchange EX_ID --position percentage --amount 3 --dynamic-leverage
```

Replacements inherit the pruned bot's `exchangeId`, `position`, `amount`,
and leverage mode so the agent can copy-paste straight into `sf-deploy`.