name: memory-ops description: > Unified memory interface with associative search and OMC integration. Triggers on "기억해", "저장해", "메모리", "뭐였지", "지난번에", "기록해", "잊지마", "정리해", "요약해줘", "remember", "save", "store", "what was it", "last time", "record this", "don't forget", "organize", "summarize" and similar. v3: Default to associative search mode (5-signal ranking). Uses memory-specialist agent for complex operations. Auto-indexes trigrams via PostToolUse hook.

Memory Operations — Persistent Memory Management Skill

Preamble

Before executing this skill:

Load context from memory:

memory_search(query: "{skill-relevant-query}", associative: true, limit: 5)
memory_search(tag: "{skill-name}", limit: 3)

Review returned memories for relevant past context, decisions, and patterns.

Check OMC state for active work:
```
state_get_status()
```
If conflicting active tasks exist, warn the user before proceeding.

Detect current branch (for git-related skills):

git rev-parse --abbrev-ref HEAD 2>/dev/null || echo "not-a-git-repo"

Check proactive mode:
```
state_read("occ-proactive")
```
If "false": do NOT proactively suggest other OpenClaw-CC skills during this session. Only run skills the user explicitly invokes.

Log skill activation:

memory_daily_log(type: "note", entry: "Skill activated: /{skill-name}")

Memory Context Loading

Before starting work, load relevant context from the 3-layer memory system:

# Search for related past work
memory_search(query: "{task description}", associative: true, limit: 5)

# Search by relevant tags
memory_search(tag: "{relevant-tag}", limit: 3)

# Check for recent related daily logs
memory_search_date(start: "{7 days ago}", end: "{today}", category: "daily-logs", limit: 5)

Use retrieved context to:

Avoid repeating past mistakes
Build on previous decisions
Maintain consistency with established patterns
Reference related knowledge graph nodes

If critical related memories exist, summarize them before proceeding:

Found {N} related memories:
- {memory_1 title}: {brief relevance}
- {memory_2 title}: {brief relevance}

Core Principles

Store all important information: Decisions, insights, and plans from conversations go into memory
Store with structure: Assign appropriate category, tags, and importance level
Retrieve when needed: Run memory_search first whenever related context is required
Periodically organize: Clean up old/duplicate memories and consolidate summaries

Category Guide

Category	Purpose	Subcategory Examples
`projects`	Everything project-related	rutile, sapiens, nexus
`people`	Person info, relationships, meeting notes	ken-huang, investors
`knowledge`	Learned knowledge, technical docs	web3-security, ai-agents
`tasks`	To-dos, work in progress	urgent, backlog
`daily-logs`	Daily activity logs	(date auto-generated)
`inbox`	Unsorted temporary storage	(classify later)

Auto-Save Triggers

When the following patterns appear in conversation, automatically suggest saving to memory:

Decision: "we decided to ~", "going with ~", "direction is ~"
Plan/Strategy: "planning to ~ going forward", "the plan is ~"
Person info: New name + role/relationship
Insight: "realized that", "the key thing is", "most important is"
Data/Metrics: Revenue, deadlines, prices, or other specific figures
Risk/Issue: Problems, risks, cautions

Store Workflows

1. Quick Store

User: "Remember this — Ken Huang is co-chair of CSA and we're meeting in SF on 3/25"

→ memory_store:
  category: "people"
  subcategory: null
  title: "Ken Huang - CSA Co-Chair"
  content: "Ken Huang is co-chair of the Cloud Security Alliance. Meeting in SF on 3/25."
  tags: ["ken-huang", "csa", "sf-trip", "meeting"]
  importance: 7

2. Project Context Store

User: "We decided to migrate Sapiens to PydanticAI"

→ memory_store:
  category: "projects"
  subcategory: "sapiens"
  title: "Framework Migration Decision - PydanticAI"
  content: "Decision to migrate from LangGraph to PydanticAI. Reason: ..."
  tags: ["sapiens", "migration", "pydanticai", "decision"]
  importance: 8

3. Knowledge Store

User: "Save this vulnerability pattern"

→ memory_store:
  category: "knowledge"
  subcategory: "web3-security"
  title: "Reentrancy Attack Pattern - Cross-function"
  content: [detailed content]
  tags: ["vulnerability", "reentrancy", "solidity"]
  importance: 6

Search Workflows

Proactive Context Search

Proactively search relevant memories before starting a task:

User: "Tell me the status of RSA Conference prep"

→ 1. memory_search(query: "RSA Conference")
→ 2. memory_search(tag: "rsa")
→ 3. memory_search(category: "projects", query: "rutile conference")
→ 4. Synthesize results and respond

Time-Based Lookup

User: "What did I do last week?"

→ memory_search(category: "daily-logs")
→ Summarize logs from the past 7 days

Memory Organization Workflows

Inbox Classification

memory_list(directory: "inbox")
→ Determine appropriate category for each item
→ Move via memory_update (or file move + re-index)

Duplicate Consolidation

Discover similar items with memory_search
→ Merge content
→ Delete old entries

Project Summary Refresh

memory_search(category: "projects", subcategory: "rutile")
→ Consolidate recent changes
→ Update _meta.md summary file

Daily Log Pattern

Record daily logs automatically or on user request:

memory_daily_log:
  entry: "AhnLab PoC meeting completed, received additional feature requests"
  type: "meeting"

Entry types: note, decision, todo, done, idea, meeting

Memory Tree Navigation

When user requests "show me the memory structure":

memory_list(directory: "", max_depth: 2)
→ Visualize and display the full tree structure

Importance Guide

Score	Criteria	Examples
9-10	Business critical, legal obligation	Investment contract terms, incorporation decision
7-8	Core decisions, key relationships	Technical architecture decision, key partner info
5-6	Useful reference information	Technical notes, meeting summaries
3-4	General records	Daily logs, idea memos
1-2	Temporary/one-off	Quick notes, information that will soon be obsolete

Completion Status Protocol

Every skill must end with one of these status codes:

Code	Meaning	When to Use
DONE	All steps completed, evidence provided	Root cause found + fix verified, PR created, review finished
DONE_WITH_CONCERNS	Completed with warnings or caveats	Tests pass but coverage dropped, fix applied but can't fully verify
BLOCKED	Cannot proceed, requires user intervention	3 failed attempts, missing permissions, external dependency down
NEEDS_CONTEXT	Missing information to continue	Unclear requirements, need user clarification

Escalation Rules

3-strike rule: After 3 failed attempts at any step, STOP and escalate to user. Do not continue guessing. Present what was tried and ask for direction.
Scope escalation: If fix/change touches 5+ files unexpectedly, pause and confirm with the user before proceeding.
Security uncertainty: If you are unsure about a security implication, STOP and escalate. Never guess on security.
Verification requirement: Never claim DONE without evidence.
- "Should work" → RUN IT. Confidence is not evidence.
- "Already tested earlier" → Code changed since. Test again.
- "Trivial change" → Trivial changes break production.

Output Format

═══════════════════════════════════════
Status: {DONE | DONE_WITH_CONCERNS | BLOCKED | NEEDS_CONTEXT}
Summary: {one-line description of outcome}
Evidence: {test output, verification results, or blocking reason}
═══════════════════════════════════════

Memory Ops