ooooooooooooooooooop
Community@ooooooooooooooooooop
A governance and lifecycle skill registry for Personal AI / DSH environments, covering multi-device sync, subagent delegation contracts, execution discipline, and reproducible research.
Agent Skills by ooooooooooooooooooop
Showing 23 vetted skills indexed across 1 GitHub repositories.
ooooooooooooooooooopyour-skill-name
Provides a reusable template for authoring new Skill units with frontmatter and contracts.
oooooooooooooooooooppublish-and-reuse
Synchronizes Personal AI infrastructure across devices using git ancestry-based direction resolution.
ooooooooooooooooooopagent-switchboard-ops
Orchestrates cross-model agent delegation and supervised execution through the agent-switchboard MCP broker.
ooooooooooooooooooopworld-model-runtime
Executes a persistent world-model loop of prediction, observation, and model revision for complex tasks.
ooooooooooooooooooopsubagent-execution-governance
Governs implementation subagents with contract snapshots, bounded reads, budgets, and structured BLOCKED protocols.
ooooooooooooooooooopfind-session
Searches DSH session history by title or content and reads messages and token statistics.
ooooooooooooooooooopdecision-gates
Enforces checkpoint evidence anchors and zero-token validation gates for long-running agent tasks.
ooooooooooooooooooopsimulate-elite-experts
Simulates multi-round expert panel deliberation for complex high-stakes decisions.
oooooooooooooooooooppersonal-ai-operations-review
Performs read-only operations reviews of Personal AI and DSH infrastructure health.
oooooooooooooooooooptask-mode-router
Classifies tasks by scale and risk to select the appropriate execution depth.
ooooooooooooooooooopdsh-config-sync
Export and restore DSH user configuration across devices with SHA-256 verification and credential exclusion.
ooooooooooooooooooopminimal-implementation
Constrains code changes to the minimal correct scope with verifiable evidence.
ooooooooooooooooooopclarify-before-change
Clarifies scope, risks, assumptions, and acceptance criteria before modifying code or rules.
ooooooooooooooooooopresearch-lab
Designs and executes reproducible controlled experiments via a Research Core runtime.
ooooooooooooooooooopskill-repository-maintainer
Audit, validate, and safely synchronize Codex Skill repositories across devices.
ooooooooooooooooooopenvironment-bootstrap
Restores registered Codex skills to devices with SHA-256 verification and read-only audits.
ooooooooooooooooooopchatgpt-web-bridge
Connects local agents to ChatGPT web sessions via MCP and REST bridges.
ooooooooooooooooooopskill-quality-gate
Evaluates Skill packages for trigger boundaries, output contracts, and regression quality.
ooooooooooooooooooopexecution-discipline
Enforces ten execution rules preventing polling, thrashing, and gate bypassing in long agent sessions.
ooooooooooooooooooopunified-taskflow
Manages complex multi-phase tasks with anchors, checkpoints, and acceptance tracking.
ooooooooooooooooooopnatural-rewrite
Rewrites messages, emails, and comments naturally while preserving facts, tone, and meaning.
ooooooooooooooooooopevolution-proposal
Converts audit anomaly entries into reviewable rule, skill, and plugin improvement proposals.
ooooooooooooooooooopsystematic-optimization
Guides systematic problem-solving from quantified baselines through root-cause analysis to verified structural fixes.
Frequently Asked Questions About ooooooooooooooooooop
FAQPage SchemaWhat tasks can I accomplish with ooooooooooooooooooop's skills?βΌ
You can sync Personal AI and DSH configurations across devices, govern subagent delegation with contracts and budgets, enforce execution discipline in long sessions, audit and publish Codex skill repositories, run reproducible comparison experiments, and rewrite messages naturally in Chinese or English.
Who are these skills designed for?βΌ
They target operators of Personal AI / DSH (DeepSeek Harness) infrastructure who manage multi-device environments, supervise multi-model agents (Codex, Claude, Gemini via agent-switchboard MCP), and need auditable, recoverable long-horizon task execution rather than one-off edits.
How do the environment sync and restore skills work in practice?βΌ
dsh-config-sync packages ~/.dsh settings into a sanitized, credential-free skeleton; environment-bootstrap restores registered skills to other devices with read-only audit, explicit apply, and SHA-256 verification without deleting extra target files. publish-and-reuse orchestrates pull, merge, and fresh-restore decisions automatically.
What prerequisites and dependencies do these skills require?βΌ
Core governance skills depend on task-mode-router, minimal-implementation, and execution-discipline. Sync flows require skill-repository-maintainer, dsh-config-sync, and environment-bootstrap. Cross-model delegation requires an agent-switchboard MCP channel or a native subagent channel; evolution-proposal additionally requires systematic-optimization.
How do the skills prevent runaway subagent sessions and token waste?βΌ
subagent-execution-governance replaces round-based limits with action thresholds, contract snapshots, bounded three-level reads, and token/tool budgets. execution-discipline bans no-wait polling and context bloat, while decision-gates adds evidence anchors and adversarial audits to keep long tasks aligned.