Skip to content

keithhegit/ultra-orchestration / skills-vnext/ultra-execution-control

ultra-execution-control

Verified

Dispatch and execution control for Codex Ultra vNext. Use after planning when work packages must be assigned, retries bounded, blockers surfaced, and write-lock safety enforced through DAG readiness, owned-path checks, careful or freeze guardrails, and host-driven ledger updates.

FileKindSize
SKILL.mdskill md1.5 KB
agents/openai.yamlconfig220 B
references/execution-rules.mdmarkdown464 B
---
name: ultra-execution-control
description: Dispatch and execution control for Codex Ultra vNext. Use after planning when work packages must be assigned, retries bounded, blockers surfaced, and write-lock safety enforced through DAG readiness, owned-path checks, careful or freeze guardrails, and host-driven ledger updates.
---

# Ultra Execution Control

Move work packages through execution without losing safety, ownership, or
auditability.

## Dispatch Rules

Dispatch a work package only when:

- dependencies are satisfied
- acceptance checks are concrete
- write ownership is explicit
- the task is not blocked by another active writer

If a task shares `owned_paths` with another in-flight task, serialize it.

## Ledger Rules

The ledger is the system of record. Keep it lightweight and host-driven.

- the model may propose state updates
- the host should apply the actual ledger change
- every blocker should include a short summary, evidence, and next-step ask
- every retry should increment a bounded counter

## Guardrails

Use:

- `careful` for risky or destructive commands
- `freeze` to limit writes to the declared path boundary
- `guard` when both are needed together

## Result Contract

Require every child execution unit to return:

- `task_id`
- `status`
- `summary`
- `changed_paths`
- `artifacts`
- `evidence`
- `risks`
- `followups`

If evidence is missing, do not mark the package review-ready.

Read [execution-rules](references/execution-rules.md) before dispatching.
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall