child-agent-ops
Manage child agents with standardized workflows
Multi-agent projects often face coordination chaos. This skill provides clear role definitions, checklists, and templates so leader, implementer, and reviewer agents can collaborate efficiently without confusion.
Download the skill ZIP
Upload in Claude
Go to Settings → Capabilities → Skills → Upload skill
Toggle on and start using
Test it
Using "child-agent-ops". Launch a codex child agent to implement user authentication
Expected outcome:
- [Child Agent Launch]
- - Type: Implementation
- - Task ID: AUTH-001
- - Scope: User login/logout, session management, password reset
- - Completion: User authentication works with test coverage >80%
- - Estimated time: 2 hours
- - Report checkpoint: Every 30 minutes
Using "child-agent-ops". Report status for the database migration task
Expected outcome:
- [Task Status Report]
- - Task ID: DB-042
- - Current: Schema migration complete, data validation in progress
- - Next: Rollback script review, production deployment
- - Issues: Minor performance degradation on staging (investigating)
Using "child-agent-ops". Request review for security configuration changes
Expected outcome:
- [Review Request]
- - Risk Level: HIGH
- - Changes: firewall rules, IAM policy updates
- - High-risk operations: Production access policy modifications
- - Attached: Security impact analysis, test results
Security Audit
SafeThis is a pure documentation skill containing only markdown guidelines for child agent operations. All 20 static findings are FALSE POSITIVES: hash values were misidentified as cryptographic weaknesses, Japanese text was flagged as crypto algorithms, and documentation templates with code examples triggered shell execution warnings. No executable code, scripts, network operations, filesystem access, or command execution exists. The skill provides only templates and workflow guidance.
Risk Factors
🌐 Network access (1)
⚙️ External commands (7)
Quality Score
What You Can Build
Coordinate multi-agent teams
Break down complex projects into parallel tasks, assign specialized agents, and monitor progress through standardized checkpoints and reporting.
Execute delegated tasks
Follow defined scope and completion criteria, report risks immediately, and coordinate with reviewers for code changes.
Ensure safe operations
Enforce pre-approval for high-risk operations like security changes, data deletion, or billing modifications.
Try These Prompts
Launch a codex child agent for [TASK-NAME]. Purpose: [one-line goal]. Completion criteria: [specific measurable outcome]. Estimated time: [duration]. Scope: [boundaries]. Report progress every checkpoint.
Report current status for [TASK-NAME]. Current state: [what is done]. Next: [what remains]. Issues: [blockers or risks]. Attach codex launch logs and execution logs.
Request review from a reviewer agent. Changes: [files modified]. Risk level: [low/medium/high]. High-risk operations: [none or describe]. Attach implementation summary and testing notes.
BLOCKER: Cannot proceed with [TASK-NAME] due to [reason]. Options considered: [what was tried]. Recommendation: [suggested path forward]. Awiting leader decision before continuing.
Best Practices
- Always document purpose and completion criteria before launching child agents
- Report risks and blockers immediately rather than waiting for completion
- Use codex launch logs with task IDs to track agent activities
- Require reviewer agent approval for any code changes before merging
Avoid
- Launching child agents without clear completion criteria
- Waiting until task completion to report blockers or issues
- Skipping reviewer checkpoints for security-sensitive changes
- Allowing parallel updates without leader coordination