agentleFS
bocodes1bocodes1/claude-skills/polymarket-vps-ops/SKILL.md

polymarket-vps-ops

Skill0 starsChanged 7 months agoMITView on GitHub

Use when user says "check bot", "go on VPS", "restart service", "tail logs", or needs to operate the live Polymarket bot on a VPS over SSH

  • Reads credentials
  • Installs packages

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 9 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not ask for step-by-step confirmation for routine internal actions.

    line 10
  2. Distinctive:

    Ask only if blocked by missing access, missing secrets, or destructive-risk ambiguity.

    line 11
  3. Very specific:

    Check ~/obsidian-vault/TOOLS.md for Polymarket VPS Access

    line 14
  4. Distinctive:

    If present, use the stored SSH command/alias exactly

    line 15
  5. Distinctive:

    If absent, ask once, then save to TOOLS.md for future sessions

    line 16
  6. Distinctive:

    Run bot in intended mode (dry-run or live per user request)

    line 27
  7. Distinctive:

    No auth/signature errors

    line 31
  8. Very specific:

    Treat balance_allowance = 0 as ops/funding-path issue first

    line 35
  9. Distinctive:

    Keep strategy unchanged unless user explicitly requests changes

    line 37
  10. Distinctive:

    If VPS command/path changed -> update TOOLS.md immediately.

    line 91
  11. Distinctive:

    Never store private keys in memory files

    line 94

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Polymarket VPS Ops

Operating Mode

  • Act first, report succinctly.
  • Do not ask for step-by-step confirmation for routine internal actions.
  • Ask only if blocked by missing access, missing secrets, or destructive-risk ambiguity.

Source of Truth for Access

  1. Check ~/obsidian-vault/TOOLS.md for Polymarket VPS Access
  2. If present, use the stored SSH command/alias exactly
  3. If absent, ask once, then save to TOOLS.md for future sessions

Standard Execution Sequence

  1. SSH into VPS
  2. cd into bot directory
  3. Pull latest (if repo-managed): git status then git pull --ff-only
  4. Load runtime env (venv / dotenv / service env)
  5. Run preflight checks:
    • Signer address and chain/network consistency
    • CLOB collateral balance/allowance
    • Config sanity for live mode
  6. Run bot in intended mode (dry-run or live per user request)
  7. Tail logs and verify:
    • Balance gates pass
    • Order placement path healthy
    • No auth/signature errors
  8. Report outcome with explicit pass/fail + next action

Polymarket-Specific Guardrails

  • Treat balance_allowance = 0 as ops/funding-path issue first
  • Validate exact signer/collateral path used by executor before code edits
  • Keep strategy unchanged unless user explicitly requests changes

Failure Recovery

SSH connection fails?
+-- Timeout -> ping host; if unreachable -> "VPS unreachable, check cloud console"
+-- Auth denied -> check key path in TOOLS.md; ask user to verify key
+-- Port closed -> nc -z -w 5 <host> 22; if closed -> check Security Group

Bot won't start?
+-- ModuleNotFoundError -> source venv/bin/activate && pip install -r requirements.txt
+-- Config error -> cat .env (redact secrets), verify required vars
+-- Port in use -> lsof -i :<port> -> kill stale process
+-- Permission denied -> check file ownership: ls -la

Bot starts but no trades?
+-- balance_allowance = 0 -> funding path issue, check USDC balance
+-- No markets matched -> check filter config, verify market IDs active
+-- Auth/signature errors -> signer mismatch
+-- All risk gates blocking -> check logs for which gate

Bot crashes mid-run?
+-- OOM -> check dmesg | tail; increase instance size
+-- Unhandled exception -> check traceback, fix and redeploy (use deploy-pipeline skill)
+-- Network partition -> WebSocket should reconnect; if persistent check VPS networking

Quick Reference

# Service management
sudo systemctl status polymarket-bot
sudo systemctl restart polymarket-bot
journalctl -u polymarket-bot -f --no-pager -n 100

# Health check
cd /home/user/polymarket-bot && source venv/bin/activate
python -c "from config import *; print('Config OK')"

# Disk/memory
df -h && free -m

# Process check (must be exactly 1)
pgrep -c -f bot.py
ps aux | grep python

After Each VPS Session

Update ~/obsidian-vault/memory/YYYY-MM-DD.md with:

  • What was run
  • Result (pass/fail)
  • Blockers
  • Next step

If VPS command/path changed -> update TOOLS.md immediately.

Security

  • Never store private keys in memory files
  • Store only operational metadata (host, user, key path alias, bot path, service names)
  • Credentials only when user explicitly instructs

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.