task-execution-orchestration
Skill2 starsChanged 4 months agoMITView on GitHub
Orchestrate sub-agents for task execution and validate work via validation sub-agents. Use when: executing planned features, managing multi-agent workflows, or validating completed work against acceptance criteria.
In this repository6 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- design-taste-frontendSkill
- feature-planningSkill
- frontend-designSkill
- web-design-guidelinesSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 49
Check for any validation failures
- Very specific: line 89
Never directly implement code or make changes yourself - always delegate to sub-agents via runSubagent
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Task Execution & Orchestration Skill
Overview
This skill guides the execution of planned features by orchestrating sub-agents for task execution and validating the work via dedicated validation sub-agents.
CRITICAL INPUT FORMAT: This skill expects a JSON manifest from the feature-planning skill. The input should be a JSON structure containing feature details, tickets, subtasks, acceptance criteria, and validation requirements.
CRITICAL EXECUTION METHOD: You MUST use the runSubagent tool to execute subtasks. Do not directly implement code or make changes yourself. Always delegate task execution to appropriate sub-agents using the runSubagent tool, and use validation sub-agents to verify completed work against acceptance criteria.
Step-by-Step Process
1. Orchestration Plan Review
Input: JSON planning manifest from the feature-planning skill
Actions:
- Parse the JSON planning manifest to understand the tickets and subtasks
- Identify the execution order based on dependencies
- Determine which tasks can be executed in parallel
Output: Execution plan with task ordering
2. Sub-Agent Assignment & Execution
For each subtask in the planning manifest JSON:
- MUST use the
runSubagenttool to assign the appropriate sub-agent for execution based on the task type - Provide the sub-agent with clear instructions and acceptance criteria via the
runSubagenttool's prompt parameter - Monitor the execution progress through the sub-agent's response
Execution Order:
- Sequential dependencies (what must run first)
- Parallelizable tasks (can be executed simultaneously using multiple
runSubagentcalls)
3. Validation Sub-Agent Assignment
For tasks requiring validation:
- Assign a validation sub-agent to verify the work
- Provide the validation sub-agent with the acceptance criteria and validation criteria from the JSON manifest
- Check for any validation failures
4. Orchestration Manifest Tracking
Track the status of each ticket and subtask using the following JSON structure:
{
"feature": "<Feature Name>",
"tickets": [
{
"id": "TICKET-001",
"title": "<Ticket Title>",
"status": "in-progress | completed | failed",
"subtasks": [
{
"id": "SUB-001-1",
"title": "<Subtask Title>",
"assignedAgent": "sub-agent",
"validationAgent": "validation-sub-agent or null",
"status": "not-started | in-progress | completed | failed",
"validationStatus": "pending | passed | failed"
}
]
}
]
}Decision Points & Branching Logic
When to Retry Failed Tasks
- Retry: For transient errors or incorrect implementation that can be fixed
- Re-plan: For fundamental misunderstandings of requirements or architectural issues
Agent Assignment Rules
- MUST use the
runSubagenttool for all task execution based on the nature of the work - MUST use the
runSubagenttool with validation sub-agents to verify completed work against acceptance criteria - Never directly implement code or make changes yourself - always delegate to sub-agents via
runSubagent
Quality Criteria & Completion Checks
For Orchestration Phase
- Execution graph correctly represents sequential and parallel tasks
- Each subtask has an assigned execution agent
- Validation requirements are specified for each task
For Validation Phase
- Validation sub-agents have clear acceptance criteria from the JSON manifest
- All validation failures are tracked and can trigger re-execution
- Final feature integration is validated end-to-end
Example Prompts to Try This Skill
- "Execute the planned face swap feature tickets using sub-agents and validate the work"
- "Orchestrate the execution of the queue/batch processing tickets with validation steps"
- "Execute the MCP server integration subtasks and validate against acceptance criteria"
Related Skills
- feature-planning: For planning the features and breaking them down into tickets with subtasks in JSON format
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

