agent-monitor

Detailed monitoring of Pi agents in WezTerm. Shows full task, recent activity, and last tool output for each agent.

16 stars

Best use case

agent-monitor is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

Detailed monitoring of Pi agents in WezTerm. Shows full task, recent activity, and last tool output for each agent.

Teams using agent-monitor should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

  • You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

  • You only need a quick one-off answer and do not need a reusable workflow.
  • You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/agent-monitor/SKILL.md --create-dirs "https://raw.githubusercontent.com/diegosouzapw/awesome-omni-skill/main/skills/data-ai/agent-monitor/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/agent-monitor/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How agent-monitor Compares

Feature / Agentagent-monitorStandard Approach
Platform SupportNot specifiedLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

Detailed monitoring of Pi agents in WezTerm. Shows full task, recent activity, and last tool output for each agent.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

SKILL.md Source

# Agent Monitor

Get detailed status of each Pi agent.

## Run Detailed Status Check

```bash
echo "🧠 DETAILED AGENT STATUS - $(date '+%H:%M:%S')"
echo ""

PANES=$(bb wezterm list 2>/dev/null)
if [ $? -ne 0 ] || [ -z "$PANES" ]; then
  echo "⚠️ WezTerm not accessible"
  exit 0
fi

echo "$PANES" | jq -r '.data[] | "\(.pane_id)|\(.title)|\(.cwd)"' | while IFS='|' read -r PANE_ID TITLE CWD; do
  CWD_CLEAN=$(echo "$CWD" | sed 's|file://||')
  PROJECT=$(echo "$CWD_CLEAN" | xargs basename 2>/dev/null || echo "unknown")
  
  [[ "$TITLE" != *"π"* ]] && continue
  
  SESSION_PATH=$(echo "$CWD_CLEAN" | sed 's|/|-|g' | sed 's|^-||')
  SESSION=$(ls -t "$HOME/.pi/agent/sessions/--${SESSION_PATH}--"/*.jsonl 2>/dev/null | head -1)
  [ -z "$SESSION" ] && continue
  
  STOP=$(tail -1 "$SESSION" | jq -r '.message.stopReason // "null"')
  ROLE=$(tail -1 "$SESSION" | jq -r '.message.role // "unknown"')
  [ "$STOP" = "stop" ] && [ "$ROLE" = "assistant" ] && STATUS="✅ IDLE" || STATUS="🔄 WORKING"
  
  echo "═══════════════════════════════════════════════════════════════"
  echo "## Pane $PANE_ID - $PROJECT - $STATUS"
  echo ""
  
  echo "**Last user request:**"
  tail -100 "$SESSION" | jq -r 'select(.type=="message" and .message.role=="user") | .message.content | if type=="array" then .[0].text else . end' 2>/dev/null | tail -1
  echo ""
  
  echo "**Recent activity (last 5 messages):**"
  tail -50 "$SESSION" | jq -r 'select(.type=="message") | "\(.message.role): \(.message.content | if type=="array" then ([.[] | if .type=="text" then .text[:150] elif .type=="toolCall" then "[\(.name)]" else .type end] | join(" ")) else .[:150] end)"' 2>/dev/null | tail -5
  echo ""
  
  echo "**Last tool output (truncated):**"
  tail -20 "$SESSION" | jq -r 'select(.type=="message" and .message.role=="toolResult") | .message.content | if type=="array" then .[0].text else . end' 2>/dev/null | tail -1 | head -c 300
  echo ""
  echo ""
done

echo "═══════════════════════════════════════════════════════════════"
```

## Output Explained

For each agent shows:
- **Pane + Project + Status** (idle/working)
- **Last user request** - the full task they were given
- **Recent activity** - last 5 messages (role + truncated content/tool calls)
- **Last tool output** - what the last command returned

## After Review

Analyze each agent and decide:
1. **Is it stuck?** Same output for too long → send "status?"
2. **Is it done?** ✅ IDLE → assign new task
3. **Is it duplicated?** Multiple panes same work → stop extras
4. **Does it need help?** Errors in output → intervene

## Actions

```bash
# Send task
bb wezterm send <pane_id> "your task"

# Check full session
tail -100 <session_file> | jq '.message'

# Interrupt
bb wezterm send <pane_id> "stop, summarize progress"
```

Related Skills

monitor-db

16
from diegosouzapw/awesome-omni-skill

Workflow for monitor-db

allegro-monitor

16
from diegosouzapw/awesome-omni-skill

Monitor Allegro.pl prices and get alerts when items drop below your threshold.

Data Quality Monitoring

16
from diegosouzapw/awesome-omni-skill

Data Quality (DQ) Monitoring is the continuous process of validating data against predefined rules and expectations. In a modern data stack, monitoring must happen at every stage: **Ingestion**, **Tra

azure-monitor-opentelemetry-ts

16
from diegosouzapw/awesome-omni-skill

Instrument applications with Azure Monitor and OpenTelemetry for JavaScript (@azure/monitor-opentelemetry). Use when adding distributed tracing, metrics, and logs to Node.js applications with Appli...

azure-monitor-ingestion-py

16
from diegosouzapw/awesome-omni-skill

Azure Monitor Ingestion SDK for Python. Use for sending custom logs to Log Analytics workspace via Logs Ingestion API.

apify-brand-reputation-monitoring

16
from diegosouzapw/awesome-omni-skill

Track reviews, ratings, sentiment, and brand mentions across Google Maps, Booking.com, TripAdvisor, Facebook, Instagram, YouTube, and TikTok. Use when user asks to monitor brand reputation, analyze...

agent-performance-monitor

16
from diegosouzapw/awesome-omni-skill

Expert performance monitor specializing in system-wide metrics collection, analysis, and optimization. Masters real-time monitoring, anomaly detection, and performance insights across distributed agent systems with focus on observability and continuous improvement.

bgo

10
from diegosouzapw/awesome-omni-skill

Automates the complete Blender build-go workflow, from building and packaging your extension/add-on to removing old versions, installing, enabling, and launching Blender for quick testing and iteration.

Coding & Development

partner-revenue-desk

16
from diegosouzapw/awesome-omni-skill

Operating model for tracking, attributing, and accelerating partner-sourced revenue.

parallel-data-enrichment

16
from diegosouzapw/awesome-omni-skill

Structured company and entity data enrichment using Parallel AI Task API with core/base processors. Returns typed JSON output. No binary install — requires PARALLEL_API_KEY in .env.local.

parallel-agents

16
from diegosouzapw/awesome-omni-skill

Multi-agent orchestration patterns. Use when multiple independent tasks can run with different domain expertise or when comprehensive analysis requires multiple perspectives.

paper-writing-assistant

16
from diegosouzapw/awesome-omni-skill

Assist in drafting research papers and meeting notes, enforcing academic rigor and formatting.