Automated summary
Monitors an Agent Mail inbox to surface pending messages, unresolved help requests, and recent completions across cooperating agents.
Strengths
- Provides a clear step-by-step workflow including availability checks, agent identity resolution, and project key lookup.
- Categorizes messages into pending items, HELP_REQUESTs, and completions, prioritizing blocked agents.
- Includes a --watch mode for continuous polling during orchestration sessions.
- Falls back to an HTTP endpoint when Agent Mail MCP tools are unavailable.
Limitations
- Requires Agent Mail through MCP or a local HTTP endpoint at localhost:8765, and relies on subject conventions such as HELP_REQUEST/OFFERING_READY.
- Output formatting, watch-mode behavior, and transport details are deferred to external reference files, so the SKILL.md is not fully self-contained.
- Contains an imperative execution directive ('YOU MUST EXECUTE THIS WORKFLOW') that should be reviewed when skills are treated as untrusted content.
Use it when an agent needs a quick check for pending messages, help requests, or task-completion notices in a multi-agent project.
Avoid it when Agent Mail is not deployed or the project does not use this messaging system for agent coordination.
Security analysis
CautionThe skill is a read-oriented inbox monitor. Its Bash usage is limited to a localhost health check, git rev-parse, and hostname, and it only queries a local Agent Mail service via MCP or localhost HTTP. There is no destructive command, no curl|sh, no base64 payload, no secret/token exfiltration, and no attempt to disable safety. It is therefore 'caution' only because it invokes bash/network tools for legitimate coordination purposes, not because of any malicious behavior.
- •Uses Bash to run curl against http://localhost:8765 (local health check / possible HTTP fallback) — network access is local-only but still an external call.
- •Reads agent identity from environment variables (OLYMPUS_DEMIGOD_ID, AGENT_NAME, USER) and falls back to USER-hostname; minor local info exposure in output, but nothing is transmitted off-host.
- •Requires an external Agent Mail MCP server / localhost HTTP service whose behavior is outside this file's control.
- •Uses Bash plus a network call (curl http://localhost:8765/health) — the endpoint is loopback-only, so exposure is minimal, but the skill does assume a local service and shell execution.
- •Reads environment variables (OLYMPUS_DEMIGOD_ID, AGENT_NAME, USER) and hostname/git paths to derive identity; values are only echoed/used for MCP parameters, not transmitted externally.
- •Watch mode instructs continuous polling of a local message store; harmless but long-running if left unattended.
Examples
/inboxCheck my Agent Mail inbox and list any unresolved HELP_REQUESTs first, then recent completions./inbox --watchname: inbox description: 'Agent Mail inbox monitoring. Check pending messages, HELP_REQUESTs, and recent completions. Triggers: "inbox", "check mail", "any messages", "show inbox", "pending messages", "who needs help".' allowed-tools: Read, Grep, Glob, Bash model: haiku metadata: tier: session dependencies: []
Inbox Skill
Quick Ref: Monitor Agent Mail from any session. View pending messages, help requests, completions.
YOU MUST EXECUTE THIS WORKFLOW. Do not just describe it.
Monitor Agent Mail messages for coordination across agents.
Requires: MCP Agent Mail tools OR HTTP endpoint at localhost:8765.
Invocation
/inbox # Show current inbox state
/inbox --watch # Continuous polling mode
Execution Steps
Given /inbox [--watch]:
Step 1: Check Agent Mail Availability
# Check if Agent Mail MCP tools are available
# Look for tools starting with mcp__mcp-agent-mail__
# Alternatively, check HTTP endpoint
curl -s http://localhost:8765/health 2>/dev/null && echo "Agent Mail HTTP available" || echo "Agent Mail not running"
Step 2: Determine Agent Identity
# Check environment for agent identity
if [ -n "$OLYMPUS_DEMIGOD_ID" ]; then
AGENT_NAME="$OLYMPUS_DEMIGOD_ID"
elif [ -n "$AGENT_NAME" ]; then
AGENT_NAME="$AGENT_NAME"
else
# Default to asking or using hostname
AGENT_NAME="${USER:-unknown}-$(hostname -s 2>/dev/null || echo local)"
fi
# Get project key (current repo path)
PROJECT_KEY=$(git rev-parse --show-toplevel 2>/dev/null || pwd)
echo "Agent: $AGENT_NAME"
echo "Project: $PROJECT_KEY"
Step 3: Fetch Inbox (MCP Method)
Use MCP tool if available:
Tool: mcp__mcp-agent-mail__fetch_inbox
Parameters:
project_key: "<project-key>"
agent_name: "<agent-name>"
Parse results into categories:
- Pending: Messages without acknowledgement
- HELP_REQUEST: Messages with subject containing "HELP_REQUEST"
- Completions: Messages with subject "OFFERING_READY" or "DONE"
Step 4: Search for HELP_REQUESTs Needing Response
Search for unresolved help requests:
Tool: mcp__mcp-agent-mail__search_messages
Parameters:
project_key: "<project-key>"
query: "HELP_REQUEST"
Filter to those without HELP_RESPONSE in same thread.
Step 5: Get Recent Completions
Search for done messages:
Tool: mcp__mcp-agent-mail__search_messages
Parameters:
project_key: "<project-key>"
query: "OFFERING_READY OR DONE OR COMPLETED"
Step 6: Summarize Threads (Optional)
For active threads with multiple messages:
Tool: mcp__mcp-agent-mail__summarize_thread
Parameters:
project_key: "<project-key>"
thread_id: "<thread-id>"
Step 7: Display Results
Read references/output-format.md for the display template and example session.
Watch Mode
Read references/watch-mode.md for watch mode polling loop, alerting, and message summaries.
Transport Details
Read references/transport-reference.md for MCP tool reference, HTTP fallback, and setup instructions.
Key Rules
- Check regularly - Agents may be waiting for help
- Prioritize HELP_REQUESTs - Blocked agents waste resources
- Acknowledge completions - Closes the coordination loop
- Use watch mode - For active orchestration sessions
Examples
Checking Inbox During Active Work
User says: /inbox
What happens:
- Agent checks for MCP Agent Mail tools, finds them available
- Agent determines identity from $AGENT_NAME or hostname, gets project key from git
- Agent fetches inbox using MCP tool, finds 3 pending messages
- Agent searches for HELP_REQUESTs, finds 1 unresolved request from worker-2
- Agent gets recent completions, finds 2 OFFERING_READY messages
- Agent displays formatted results showing pending help request and recent completions
- Agent suggests: "Worker-2 needs help with database schema migration"
Result: Inbox shows 1 critical help request requiring attention and 2 completed tasks.
Watch Mode for Orchestration
User says: /inbox --watch
What happens:
- Agent enters polling loop checking inbox every 30 seconds
- Agent displays initial state: 0 pending, 0 help requests
- After 2 minutes, new HELP_REQUEST appears from worker-5
- Agent alerts user with notification and message summary
- Agent continues watching, detects OFFERING_READY from worker-3
- Agent maintains live display of inbox state until user interrupts
Result: Continuous monitoring mode catches new help requests in real-time for quick response.
Troubleshooting
| Problem | Cause | Solution |
|---------|-------|----------|
| "Agent Mail not running" error | MCP server not started or HTTP endpoint down | Start Agent Mail server: check MCP config or run standalone server on localhost:8765. Verify with curl http://localhost:8765/health. |
| Empty inbox despite active workers | Wrong agent name or project key | Verify $AGENT_NAME matches worker expectations. Check project key with git rev-parse --show-toplevel. Use absolute path, not relative. |
| HELP_REQUESTs not showing | Missing search or filter issue | Verify search query includes "HELP_REQUEST" string. Check message subjects match protocol. Use search_messages tool to debug. |
| Watch mode exits immediately | Polling error or missing dependency | Check MCP tools work individually first. Verify --watch flag parsing. Fall back to manual polling: run /inbox repeatedly instead of --watch. |
References
- Agent Mail Protocol: See
skills/shared/agent-mail-protocol.mdfor message format specifications - Parser (Go):
cli/internal/agentmail/- shared parser for all message types
From the file to a first result
The download contains exactly the content displayed on this page. This copy may differ from the latest version in the source repository.
View the source and supporting filesThis download contains SKILL.md only. Any scripts, references, templates and dependencies are not included. Get the required files from the source, preserving their directory structure, before continuing.
1. Download the file
Download SKILL.mdFrom the root of a test project, save the file as .skill-downloads/inbox-monitoring-msba3mj4/SKILL.md, or use the command below. This folder is for inspecting the skill before installation.
Command for macOS/Linux or WSL: requires a POSIX shell and curl. On Windows without WSL, use the download button and create the folder manually.
mkdir -p '.skill-downloads/inbox-monitoring-msba3mj4' && curl --fail --show-error --location 'https://skills-guides.com/api/skills/inbox-monitoring-msba3mj4/download' --output '.skill-downloads/inbox-monitoring-msba3mj4/SKILL.md.part' && mv '.skill-downloads/inbox-monitoring-msba3mj4/SKILL.md.part' '.skill-downloads/inbox-monitoring-msba3mj4/SKILL.md'2. Check prerequisites
Read SKILL.md before using it. If it references scripts or external services, inspect them and prepare the required access in your test environment.
Tools declared by the author (check availability in your agent): Read, Grep, Glob, Bash
No recognized prerequisites section in this file. This does not mean it is self-contained: libraries, tools and access requirements may be mentioned elsewhere in the content or source repository.
The platforms on this page are catalog indications. The paths below follow tool documentation; they do not prove this skill has been run in those tools. No minimum tool version is confirmed for this listing.
3. Load the skill
After inspection, copy the complete folder into your project using one of the commands below, then open a new tool session. The command stops if a folder with the same name already exists: compare the files before replacing it.
Claude Code
test -f '.skill-downloads/inbox-monitoring-msba3mj4/SKILL.md' && test -s '.skill-downloads/inbox-monitoring-msba3mj4/SKILL.md' && mkdir -p '.claude/skills' && test ! -e '.claude/skills/inbox' && cp -R '.skill-downloads/inbox-monitoring-msba3mj4' '.claude/skills/inbox'OpenSkills
OpenSkills must already be installed. The local install/read procedure is checked with OpenSkills 1.5.0, which requires Node.js ≥ 20.6.0. Catalog slugs are not resolved automatically.
Official OpenSkills installationopenskills install './.skill-downloads/inbox-monitoring-msba3mj4'openskills read 'inbox-monitoring-msba3mj4'Reading should display the installed file content and its base directory. It does not run the skill or install its dependencies.
For an agent that reads AGENTS.md, then sync the index. This command changes AGENTS.md: review the diff before using it.
openskills syncManual method or another tool: attach SKILL.md to a conversation, or open it in an agent that can read files. Include any required supporting resources. No universal folder is recognized by every tool.
4. Try a first task
Use a short, synthetic input: an API endpoint, a small document or an interface brief. In a new conversation, attach the file or give the agent its path, then adapt this request:
Read .skill-downloads/inbox-monitoring-msba3mj4/SKILL.md and apply its instructions to this example: [describe a test input]. Identify any missing files, dependencies or access. Produce the result requested by the skill without inventing data or running unreviewed scripts.Expected result: output matching the skill instructions, or a precise list of missing prerequisites. A read confirmation alone does not validate a first use. If an external resource is unavailable, that path remains untested.
React Hook Form Best Practices
Development
Master React Hook Form: avoid anti-patterns, manage subscriptions, and handle reset, dirty state, and controlled inputs correctly.
Emotion to Mantine Migration
Development
Migrate Emotion styled-components to Mantine components with style props and CSS modules. Use when converting .styled.tsx files or removing @emotion imports.
Scope Creep Detector
Development
Analyzes git diffs against a stated intent to detect scope creep, unrelated files, broad PRs, and oversized changes, then suggests keep, split, or justify actions.