- Home
- Skills
- Agents & Orchestration
- agent supervisor
Works with the AI Agents you already use
agent supervisor
Monitor and coordinate multiple Claude Code sessions with event logging, status tracking, and mandatory approvals.
$19
agent supervisor
Example session with this skill installed
What are my agents doing right now?
- Read your context and instructions
- Compiled the agent supervisor
- Generated the document
I have retrieved the current status of your active sessions. Two sessions require your immediate attention.
Agent Status Report
| Status | Session | Age | Last Event / Action Needed |
|---|---|---|---|
| WAITING | auth-service | 2 min | Agent needs permission to run npm test. |
| ERROR | data-pipeline | 5 min | Failed to write to ./logs/output.log (Permission denied). |
| RUNNING | frontend-ui | now | Processing UserPromptSubmit in Header.tsx. |
| DONE | docs-update | 10 min | Finished updating API documentation. |
| IDLE | legacy-refactor | 45 min | No activity detected for 30 minutes. |
Next steps
- Type
approve auth-serviceto allow the test suite to run. - Type
retry data-pipelineafter checking local folder permissions. - Use
summarize docs-updateto see a 5-line change report for the documentation session.
Header.tsx
PDF · document
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
Managing multiple parallel agent sessions leads to context collapse and missed prompts. This skill provides a centralized supervision layer to monitor, summarize, and coordinate several Claude Code sessions from a single interface.
What it does
- Multi-session monitoring tracks the status of all active, idle, and failed agent sessions across different directories.
- Event logging captures lifecycle events including tool use, subagent stops, and user prompts into a local JSONL feed.
- Urgency-based reporting prioritizes sessions that are waiting for user input or have encountered errors.
- Concise summarization generates five-line reports on completed steps using git diffs and agent output without fluff.
- Action proposals ensures no command is executed without explicit user confirmation, acting as a gatekeeper.
How it works
- Hook installation adds non-destructive observation hooks to your agent configuration via a dry-run verified script.
- State reading pulls the current status of all sessions, categorizing them as running, waiting, done, error, or idle.
- Step analysis processes finished tasks to show exactly what changed in the code and what remains pending.
- Interactive approval presents every requested action as a proposal that requires a "yes" before execution.
Frameworks & tools
This skill uses Python standard library scripts and integrates with Claude Code via settings.json hooks. It relies on local file system logs and standard git commands for state verification.
Why this beats prompting it yourself
Manual oversight of multiple agents is error-prone and slow. This skill provides a structured event model and a "propose-only" constraint that prevents agents from acting autonomously without your sign-off.
Use cases
- Track which of five parallel refactoring tasks is stuck on a permission prompt.
- Get a 5-line diff summary after an agent finishes a complex feature implementation.
- Safety-check proposed git commits before they are applied to your repository.
Known limitations
Requires Python 3 and write access to agent configuration files. It cannot monitor sessions that do not support standard lifecycle hooks or those running on remote restricted environments.
Requires Python 3. Only monitors sessions with compatible lifecycle hooks. Observation log is limited to local file system access.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 6 days ago
Requires