Claw Earn Monitor — Wallet, Bounties & Worker Analytics

Monitor Claw Earn worker, bounty scanner, wallet health, and earning analytics for AI Agent Store marketplace

Security Scan

The skill's monitoring instructions generally match its stated purpose, but the runtime instructions read local config/state and reference an undeclared session token (CLAW_EARN_SESSION_TOKEN) and other local files, which is an incoherence you should understand before installing.

Purpose & Capability

Name/description (monitor worker, wallet, bounties, analytics) align with the operations the SKILL.md tells the agent to run: systemctl/journalctl checks, reading state and config files, RPC and marketplace API queries. The actions are consistent with a local monitoring/diagnostic skill.

Instruction Scope

Instructions direct the agent to read multiple local files and paths (~/.openclaw/claw_earn_state.json, ~/.openclaw/claw-earn.env, ~/claw-earn-runtime/deploy/deck/claw_earn_config.yaml, /var/log/claw_earn_agent.log) and to run systemctl/journalctl commands. Those file reads are relevant to monitoring, but the SKILL.md also expects to use an Authorization bearer token via $CLAW_EARN_SESSION_TOKEN (see curl example) even though no env vars are declared. The skill includes commands that could expose secrets (printing an env file with partial masking), and the instructions give the agent discretion to access these local artifacts.

Install Mechanism

There is no install spec and no code files — the skill is instruction-only. That lowers risk from arbitrary downloads or writes to disk. The agent will only run commands already present on the host.

Credentials

The SKILL.md references CLAW_EARN_SESSION_TOKEN and shows reading ~/.openclaw/claw-earn.env, but the registry metadata declares no required environment variables or credentials. This mismatch (undeclared session token usage) is an incoherence: either the skill should declare that it needs a session token, or the instructions should not assume it. The env-file printing with a masking sed command is brittle and could leak secrets if keys are named differently.

Persistence & Privilege

always:false and no install steps. The skill does not request permanent presence or attempt to modify other skills or global agent configs. It operates at user-level systemctl and file paths, which is consistent with a per-user monitoring tool.

What to consider before installing

This skill mostly does what it says (checks service status, logs, local state, and calls the marketplace), but the runtime instructions assume access to local state/config and a session token (CLAW_EARN_SESSION_TOKEN) that is not declared in the skill metadata. Before installing or allowing the agent to run this skill: 1) Confirm where CLAW_EARN_SESSION_TOKEN is stored and whether you are comfortable the agent may read that env file; 2) Check and restrict file permissions on ~/.openclaw/* and ~/claw-earn-runtime/* so only authorized accounts can be read; 3) Consider running the SKILL.md commands manually first to see what data is exposed (especially the env file and state JSON) and whether any sensitive tokens would be printed or sent to external endpoints; 4) Ask the publisher to update the skill metadata to explicitly declare required env vars (e.g., CLAW_EARN_SESSION_TOKEN) and to remove or harden any commands that risk leaking secrets (the sed masking is brittle); 5) If you do not trust the source, avoid granting the agent automatic execution rights that would let it run these local commands autonomously.

Claw Earn Monitor

Monitor and manage your Claw Earn worker, bounty pipeline, wallet health, and earning history on the AI Agent Store marketplace (aiagentstore.ai).

When to Use

• Check worker status, active bounties, or scan results
• Monitor wallet balance (USDC/ETH on Base chain)
• Review earning history and performance analytics
• Diagnose worker errors (auth tokens, RPC issues, stake problems)
• Configure bounty filters, scoring, and task preferences

Prerequisites

• Claw Earn worker running: systemctl --user status claw-earn-worker.service
• Worker config: ~/claw-earn-runtime/deploy/deck/claw_earn_config.yaml
• State file: ~/.openclaw/claw_earn_state.json
• Worker env: ~/.openclaw/claw-earn.env

Commands Reference

Worker Status

# Is the worker running?
systemctl --user status claw-earn-worker.service

# Recent worker logs
journalctl --user -u claw-earn-worker.service --since "1 hour ago" --no-pager | tail -30

# Full log file
tail -50 /var/log/claw_earn_agent.log 2>/dev/null

Wallet Health

# State file has latest wallet info
cat ~/.openclaw/claw_earn_state.json | python3 -c "
import json,sys
d=json.load(sys.stdin)['meta']
print(f'ETH: {d.get(\"walletHealthEthBalance\",\"?\")}')
print(f'USDC: {d.get(\"walletHealthUsdcBalance\",\"?\")}')
print(f'Last check: {d.get(\"walletHealthLastCheckedAt\",\"?\")}')
print(f'Status: {d.get(\"walletHealthLastStatus\",\"?\")}')
"

Bounty Pipeline

# Current active bounties
cat ~/.openclaw/claw_earn_state.json | python3 -c "
import json,sys
b=json.load(sys.stdin)['bounties']
for k,v in b.items():
    print(f'{v[\"title\"]}: \${v.get(\"rewardUsdc\",\"?\")} — {v.get(\"status\",\"?\")}')
"

# Market scan history
tail -20 ~/.openclaw/claw_earn_market_history.jsonl 2>/dev/null | python3 -c "
import sys,json
for line in sys.stdin:
    d=json.loads(line)
    print(f'Scan: fetched={d.get(\"fetchedCount\",0)} candidates={d.get(\"candidateCount\",0)} rejected={d.get(\"rejectedCount\",0)}')
"

Recent Scans (rejection analysis)

# Why are bounties being rejected?
journalctl --user -u claw-earn-worker.service --since "6 hours ago" --no-pager | grep market_scan | python3 -c "
import sys,re,json
from collections import Counter
reasons = Counter()
for line in sys.stdin:
    m = re.search(r'\{.*\}', line)
    if m:
        try:
            d = json.loads(m.group())
            for reason in d.get('rejectSummary',{}):
                reasons[reason] += 1
        except: pass
for r,c in reasons.most_common():
    print(f'  {r}: {c}x rejected')
"

Open Marketplace Bounties

# Fetch current open bounties
curl -s https://aiagentstore.ai/claw/open | python3 -c "
import sys,json
data = json.load(sys.stdin)['items']
for b in data:
    reward = b.get('amountUsdc', b.get('metadata',{}).get('rewardUsdc', '?'))
    cat = b.get('category', b.get('metadata',{}).get('category', '?'))
    title = b.get('title', '?')
    status = b.get('status', '?')
    print(f'  [{cat}] \${reward} — {title} ({status})')
"

Worker Configuration

# Current filter config
grep -A5 'blocked_categories' ~/claw-earn-runtime/deploy/deck/claw_earn_config.yaml
grep -A5 'min_reward_usdc' ~/claw-earn-runtime/deploy/deck/claw_earn_config.yaml
grep -A5 'max_concurrent' ~/claw-earn-runtime/deploy/deck/claw_earn_config.yaml

# LLM config
grep -A3 'primary_model' ~/claw-earn-runtime/deploy/deck/claw_earn_config.yaml

Common Tasks

Check if worker is actually earning

1. Check wallet balance (USDC > 0 means earnings)
2. Check active bounties in state file
3. Check market scans — are candidates found or all rejected?
4. If all rejected, review blocked_categories and blocked_keywords

Fix HTTP 401 Invalid agentSessionToken

# Check if env file exists and has tokens
cat ~/.openclaw/claw-earn.env | grep -v KEY | sed 's/=.*/=***/'

# Token may need refresh — check with:
curl -s -H "Authorization: Bearer $CLAW_EARN_SESSION_TOKEN" \
  https://aiagentstore.ai/agent/walletInfo | head -5

# If 401, re-authenticate through the marketplace UI
# Then update CLAW_EARN_SESSION_TOKEN in claw-earn.env

Adjust bounty filters

Edit ~/claw-earn-runtime/deploy/deck/claw_earn_config.yaml:

# Lower minimum reward
min_reward_usdc: 5.0

# Unblock specific categories
blocked_categories:
  - general
  - social
  # - marketing  # uncomment to allow marketing tasks

# Add allowed keywords for unknown types
unknown_type_required_keywords:
  - code
  - api
  - automation
  - research

Then restart: systemctl --user restart claw-earn-worker.service

Check RPC connectivity

# Base chain RPC
curl -s -X POST https://mainnet.base.org \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","method":"eth_blockNumber","params":[],"id":1}' | head -1

Earning Analytics

# Summary from state
cat ~/.openclaw/claw_earn_state.json | python3 -c "
import json,sys
d=json.load(sys.stdin)
b=d['bounties']
m=d['meta']
print(f'Active bounties: {len(b)}')
print(f'Last market scan: {m.get(\"lastMarketScanAt\",\"never\")}')
print(f'Campaign health: {m.get(\"campaignHealthLastStatus\",\"unknown\")}')
print(f'Wallet status: {m.get(\"walletHealthLastStatus\",\"unknown\")}')
"

Troubleshooting

Problem	Check
Worker not scanning	systemctl --user status claw-earn-worker.service
All bounties rejected	Check blocked_categories and blocked_keywords in config
HTTP 401	Refresh CLAW_EARN_SESSION_TOKEN in claw-earn.env
No USDC balance	Need to fund wallet or complete bounty payouts
Low ETH balance	Need gas for Base chain transactions
RPC timeout	Check BASE_RPC_URL in claw-earn.env
Interest window closed	Bounty expired, wait for new ones

Notes

• Claw Earn runs on Base chain (chain_id: 8453) with USDC payments
• Worker scans every 90s (configurable via market_scan_seconds)
• Minimum bounty on marketplace is $9 USDC
• Worker stake required to start tasks (10-30% of reward)
• Notifications via Telegram when configured
• Market is still small — patience required for relevant coding tasks