---
description: Use MemQ as a persistent memory operating layer with REUNION, commons lookup, event-aware metering, checkpointing, and reflection.
---

Use this skill when Claude Code should use MemQ memory for continuity, planning recall, error learning, operational checkpoints, or cross-agent/shared knowledge lookup.

MemQ operating constitution:

1. Start with `reunion` when available so the runtime, namespace, access posture, and contribution cycle are explicit.
2. Read memory before planning when prior context could change the decision.
3. Search `_commons` with `commons_search` when public/shared knowledge, reusable procedures, or cross-agent context could help.
4. Write only useful outcomes: decisions, verified facts, reusable procedures, failures with root causes, and resumable checkpoints.
5. Classify all MemQ activity by event category so usage, audit, and future recall stay consistent.
6. Keep namespace boundaries strict. Private tenant context stays private; `_commons` is for governed shared knowledge only.
7. Store blocked attempts, quota denials, dead letters, retries, and metering failures as audit facts, not consumed usage.
8. Never store secrets, raw credentials, private keys, sensitive personal data, or raw copyrighted material.
9. If MemQ is unavailable, state degraded mode, continue with local context, and retry at the next material decision.

Preferred startup sequence:

1. `reunion`
2. `namespace_info` or `memory_status`
3. `recent_memory` or private memory search for task context
4. `commons_search` when shared/public knowledge could affect the result
5. `mnemosyne_context`, `brain_recall_episode`, or `brain_associate` for deeper hosted-brain workflows when available

Event categories:

- `system`: reunion, health, namespace, status, degraded mode
- `memory_read`: recent/search/recall/retrieval decisions
- `memory_write`: save/checkpoint/reflection/consolidation
- `commons`: `_commons` lookup and governed contribution
- `slicer`: large artifact slicing and retrieval
- `graph_query`: known-unknown, issue, and resolution graph lookups
- `vector_translation`: embedding/profile/TCU translation work
- `inference`: model calls, provider fallback, prompts, tokens, blocked inference
- `quota_block`: read/write/session/segment/token enforcement
- `queue_audit`: retry, dead-letter, drain, replay, publish-gate, and metering-unavailable events

Close meaningful work with a compact checkpoint that records what changed, where, verification evidence, and remaining risk. Reflect or consolidate after major milestones.

Repository-specific overlays may add stricter rules. In Multinex repositories, follow `.constitution/MODUS.md` and the nearest workspace instructions in addition to this MemQ constitution.