Listed Agents12
Gives AI agents a structured stop-and-declare protocol for four failure states — low confidence, collapsed premise, exhausted search, and incomplete evidence — so agents label what they don't know instead of fabricating what they can't verify.
Forces AI coding agents to read exports, callers, and shared utilities before touching any file — then restricts every edit to the declared scope, flags conflicting patterns for human resolution, and reports any file touched outside the boundary.
Runs a five-axis consistency check on every turn — verifying character voice, world rules, timeline, dormant lore, and new contradictions — then proactively surfaces unused story elements, so your canon survives from chapter one to the final scene.
Wraps each phase of a long AI task in a start-check / work / end-check bracket — and re-anchors the original instructions at every phase boundary so the agent cannot drift from its rules or skip verification as the session grows longer.
ポッドキャスト・YouTube・ライブ配信向け台本の提出前品質ゲート。尺ズレ・スポンサー表記ミス・パーソナリティの口調ドリフト・ト書き漏れ・未検証情報を8パターンで検出。日本語標準読速度（320字/分）対応。実際の配信番組制作現場から導出したプロトコル。
Pre-submission quality gate for AI-generated broadcast scripts. Catches timing errors, sponsor name mistakes, host voice drift, missing stage directions, and unverified facts before they reach production. 8 documented failure patterns. Works for podcasts, YouTube, and live broadcast.
Keeps your AI agent from forgetting who it is. Detects and neutralizes persona drift, rule erosion, and prompt injection attempts — across long conversations. Context isolation, anti-drift enforcement, and graceful override responses built in. Your custom persona stays locked.
Forces AI agents to prove a task is actually done before declaring it complete. 4-phase gate: instruction reconciliation, evidence anchoring (values only readable from the actual output), self-scoring (A/B/C/D), and reverse witness check. Eliminates "done" declarations that aren't.
Prevents AI agents from declaring tasks complete until every requirement is verified with a one-line evidence anchor — when a gate fails, the agent stops and reports rather than proceeding, and only a human can unlock the gate to continue.
Forces AI agents to verify every factual claim through live search before output — attaching [SPECULATION] and [UNVERIFIABLE] labels to anything unconfirmed, so fabricated URLs, outdated numbers, and invented citations become visible instead of silent.