From memory
Self-learning memory system. Orchestrates the Retrieve -> Act -> Log cycle to ensure persistent learning across sessions. Use this skill at the beginning of tasks to retrieve context and at the end to log outcomes. Triggers on starting non-trivial tasks, debugging errors, auditing memory, or storing static knowledge.
How this skill is triggered — by the user, by Claude, or both
Slash command
/memory:memory-usageThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Self-learning memory system. Orchestrates the **Retrieve -> Act -> Log** cycle to ensure persistent learning across sessions.
Self-learning memory system. Orchestrates the Retrieve -> Act -> Log cycle to ensure persistent learning across sessions.
Before starting any significant task (coding, debugging, architecture design), you MUST check for relevant past experiences.
prociq_retrieve_context with a clear description of the current goal.error_state, tools, file_patterns, or limit to focus the search.prociq_retrieve_context as the first non-trivial action, do not call prociq_list_scopes before this first retrieval.After the first prociq_retrieve_context call in a task/session:
prociq_list_scopes to discover authorized scopes.prociq_log_episode, prociq_log_note, prociq_log_episodes_batch, etc.), ensure default scope is resolved.Perform the task as requested, informed by the retrieved context.
prociq_retrieve_context again with error_state populated (and an updated task_description) to find specific solutions.prociq_log_note with the resolved default scope to store it permanently.After the task is complete (even if only partially successful or failed), you MUST record the experience.
prociq_log_episode.task_goal: Concise statement of what you tried to do.approach_taken: Summary of the steps or logic used.outcome: One of success, partial, or failure.scope: The resolved default memory scope used for this task.project or importance_hint in prociq_log_episode.prociq_log_episodes_batch.prociq_log_note for facts and knowledge that are not tied to a specific action outcome.prociq_get_memory_stats, prociq_search_episodes, and prociq_search_patterns to provide a comprehensive report.approach_taken is descriptive enough to be understood in a future session without the original conversation context.Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub prociq/memlayer-plugin --plugin memory