📖About

Ninad UAT (2026-07-13): asked Prometheus what measures Blueprint tracks — he confabulated six nonexistent services with fake telemetry. Root causes in the canonical talk-to-agent mechanism: (1) the persona is built from role/duties/personality fields but NEVER includes the agent's doctrine column — the single source of agent knowledge, so doctrine updates cannot reach conversations; (2) each agent has ONE endless thread with no reset, so a poisoned/hallucinated context anchors every later answer; (3) with no domain facts in context the model invents despite the honesty instruction. FIX (all generic, one shared mechanism per the canonical-chat policy): persona includes the doctrine section; new clear-conversation endpoint + a New-conversation button in the chat window; Prometheus's poisoned thread wiped as remediation. Regression test locks doctrine-in-persona and the reset path.

⚖️Decisions

• Elrond's bypass methodology checklist (a reminder, not a gate -- you've got this): 0. File it requested_mode='bypass' (planner-vs-bypass is a deliberate choice). bypass_start REFUSES a non-bypass initiative (CAROL-INI-1846), and the dispatcher only skips the bypass lane when the mode says bypass -- a 'planner' mistag lets Merlin's pipeline grab the placeholder step and block your finished work. 1. Filed as planned status -- let the bypass claim/activate it; never file active. 2. Open the bypass (bypass_start) with your droid id + the remediation answer (remediates_initiative_id=NNN, or remediates_nothing=True). 3. Work the blocks for your work-type: template -> design -> code -> test -> review. Do the real work; record decisions on the initiative as you make them. 4. Reality is recorded for you at close -- code (files changed), each decision, and the twin-review verdict become real activities tied to this initiative and show in the Activity Tracker like a planner run (CAROL-INI-1840). No dummy rows. 5. Keep the initiative status moving; it parks in 'reviewing' and is tagged uat-pending for you at close (CAROL-INI-1836), so the stuck-watchdog leaves it alone until UAT. 6. Close runs the gates (design/architecture compliance + caller-audit). If a gate flags something pre-existing or unrelated to your change, waive it with a clear written rationale -- audit, don't skip. 7. Bypass skips the planner's auto-orchestration, NOT the standards. Same template checklist, same review, same observability as a planner run. (elrond)
• Current state at filing (Elrond validity check): No live system facts mention any completed work for this filing. The recently closed initiatives list does not include any chat-related doctrine or reset work. Ninad's UAT observation from 2026-07-13 confirms the problem exists. (elrond)
• [status-router] planned -> executing | event=bypass_executing | bypass transition (or-bx-01)
• [status-router] executing -> reviewing | event=bypass_reviewing | bypass transition (or-bx-01)
• [status-router] reviewing -> closed | event=operator_signoff | Auto-accepted (CAROL-INI-1859): Orion-initiated, >2 days in reviewing with no objection. (el-srac-01)

✅Success criteria

• Ninad opens a fresh chat with Prometheus, asks which measures Blueprint tracks, and reads back the true answer — exactly one measure, Unblocked work share with target 8 — with zero invented service names in the reply (must_have)
• Ninad sees a New conversation button in every agent chat window; clicking it visibly empties the thread and the agent's next answers no longer reference the wiped exchange (must_have)
• When Ninad changes what an agent should know (doctrine), the agent's very next fresh chat already reflects it — no restart, no extra sync step (must_have)
• The nightly regression report shows a failing line if agent chat ever stops carrying doctrine or loses the reset button (must_have)