Skills & Prompts

MemQ skills and prompt library

Install the right MemQ skill, then use these prompt templates to make memory startup, recall, handoff, review, and checkpointing repeatable.


How to use this library

Filter by agent runtime, workflow stage, or audience. The prompts are intentionally operational: they tell the agent which MemQ tools to use, what evidence to collect, and what not to store.


Operator Library

Find the right MemQ workflow in seconds.

Search skills and prompts from one mobile-first control surface. Pick a runtime, choose a workflow stage, then hand the prompt to your agent with the exact memory tools already named.

Installable Skills

Agent runtime skills

Claude CodeOperator

MemQ Operations for Claude Code

Claude Code skill bundle for hosted MemQ MCP discovery, memory reads, memory writes, and session handoff.

CodexOperator

MemQ Operations for Codex

Reunion-first startup, task recall, Commons lookup, checkpointing, and reflection for Codex sessions.

OpenClawTeam

MemQ Operations for OpenClaw

Portable Open Skills markdown for OpenClaw-compatible agents and shared team memory workflows.

Prompt Templates

Reusable MemQ prompts

AuditReviewer

Code review with memory

Review a diff against prior bugs, product rules, and remembered deployment requirements.

search_memorycommons_searchmemory_status
Review this diff using MemQ context. Search for previous issues in the touched area, check Commons for reusable review guidance, then lead with concrete findings by severity. Verify claims against the code before reporting.
HandoffTeam lead

Cross-agent handoff

Give another agent enough context to continue without exposing irrelevant transcript noise.

save_contextadd_memoryrecent_memory
Prepare a handoff for the next agent. Include current branch, files changed, commands already run, failing or passing checks, blockers, and exact next steps. Store it in MemQ as a checkpoint and keep it concise.
RecallSolo developer

Recall a project decision

Recover prior decisions before changing code so the agent does not repeat old analysis.

search_memorybrain_recall_episode
Before editing, search MemQ for prior decisions about this feature, file path, and failure mode. If an episode is clearly relevant, recall it and extract constraints, commands, and known pitfalls.
SaveAgent operator

Save a durable implementation checkpoint

Persist what changed, what passed, and what remains so the next session resumes cleanly.

add_memorysave_contextreflect_memory
Create a MemQ checkpoint for this completed milestone. Include changed surfaces, root cause, validation commands with outcomes, deploy state, live proof, and remaining risks. Do not store secrets or private customer data.
StartAgent operator

Session boot with memory

Start a new agent session with namespace, recent memory, and Commons context established before planning.

reunionmemory_statusrecent_memorycommons_search
Start this task with MemQ. Run reunion, check memory_status, read recent_memory for this workspace, search task-relevant private memory, and use commons_search for reusable shared lessons. Summarize only the context that changes the plan.
TeamSupport

Team knowledge lookup

Use governed team memory to answer a customer or operator question without mixing private namespaces.

namespace_infosearch_memorycommons_search
Answer this using MemQ with namespace discipline. Confirm the current namespace, search private memory only for authorized customer context, use Commons for public/shared procedures, and call out when evidence is missing.