hashgraph-online/awesome-codex-plugins / plugins/beastlabai/multi-llm-plugin/skills/multi-llm
multi-llm
FlaggedOrchestrate plan reviews, task generation, implementation and code reviews across several LLM CLIs in parallel, or ask every model one free-text question about a plan. Use when the user says: review my plan, multi-llm, run a code review, generate tasks from a plan, implement this plan, ask the models about my plan.
schemas/human_decision_verdicts.schema.json
View on GitHub{
"$schema": "http://json-schema.org/draft-07/schema#",
"$id": "human_decision_verdicts.schema.json",
"title": "Human-decision judge verdict file (human_decision_batch_<n>.json)",
"description": "The file a human-decision judging subagent's reply is persisted to at output_path. The wrapper object (rather than a bare array) is what lets the ingest distinguish a truncated write from a legitimately empty verdict set. `decision` is a wire vocabulary: the ingest translates approve -> approved, skip -> skipped, salvage -> approved plus the judge_auto_decide_salvage markers.",
"type": "object",
"additionalProperties": false,
"required": ["schema_version", "verdicts"],
"properties": {
"schema_version": {"const": "1.0"},
"judge": {"type": "string", "description": "Optional: the model the judging agent says it is. Names a native (host-agent) judge in the report; an executor-run judge is named by its runner instead."},
"verdicts": {
"type": "array",
"items": {"$ref": "#/$defs/verdict"}
}
},
"$defs": {
"verdict": {
"type": "object",
"additionalProperties": false,
"required": ["group_id", "decision", "reason"],
"properties": {
"group_id": {
"type": "string",
"minLength": 1,
"description": "The stable group id (= group_hash) the verdict belongs to. Verdicts are mapped by this id, never by batch or list position."
},
"decision": {"enum": ["approve", "salvage", "skip"]},
"reason": {"type": "string"},
"salvaged_description": {
"type": ["string", "null"],
"description": "The rewritten description that gets applied. Non-null iff decision == salvage."
},
"dropped": {
"type": ["string", "null"],
"description": "What was intentionally left out. Non-null iff decision == salvage."
}
},
"allOf": [
{
"if": {"required": ["decision"], "properties": {"decision": {"const": "salvage"}}},
"then": {
"required": ["salvaged_description", "dropped"],
"properties": {
"salvaged_description": {"type": "string", "minLength": 1},
"dropped": {"type": "string", "minLength": 1}
}
}
},
{
"if": {"required": ["decision"], "properties": {"decision": {"enum": ["approve", "skip"]}}},
"then": {
"properties": {
"salvaged_description": {"type": "null"},
"dropped": {"type": "null"}
}
}
}
]
}
}
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| receiving-code-reviewobra/superpowersUse when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| requesting-code-reviewobra/superpowersUse when completing tasks, implementing major features, or before merging to verify work meets requirements | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| using-git-worktreesobra/superpowersUse when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallback | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| verification-before-completionobra/superpowersUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |