manage-memory（管理内存）

Manage Memory

Maintain Claude Code's persistent memory directory so it stays accurate, concise, and useful across sessions. MEMORY.md is loaded into the system prompt on every conversation — lines after 200 are truncated, so this file must be a lean index pointing to topic files for detail.

When to Use

• MEMORY.md is approaching the 200-line truncation threshold
• A session produced durable insights worth preserving (new patterns, architecture decisions, debugging solutions)
• A topic section in MEMORY.md has grown beyond 10-15 lines and should be extracted
• Project state has changed (renamed files, new domains, updated counts) and memory entries may be stale
• Starting a new area of work and checking whether relevant memory already exists
• Periodic maintenance between sessions to keep the memory directory healthy

Inputs

• Required: Access to the memory directory (typically ~/.claude/projects/<project-path>/memory/)
• Optional: Specific trigger (e.g., "MEMORY.md is too long," "just finished a major refactor")
• Optional: Topic to add, update, or extract

Procedure

Step 1: Assess Current State

Read MEMORY.md and list all files in the memory directory:

wc -l <memory-dir>/MEMORY.md
ls -la <memory-dir>/

Check the line count against the 200-line limit. Inventory existing topic files.

Expected: Clear picture of total lines, number of topic files, and which sections exist in MEMORY.md.

On failure: If the memory directory doesn't exist, create it. If MEMORY.md doesn't exist, create a minimal one with a # Project Memory header and a ## Topic Files section.

Step 2: Identify Stale Entries

Compare memory claims against current project state. Common staleness patterns:

1. Count drift: File counts, skill counts, domain counts that changed after additions/removals
2. Renamed paths: Files or directories that were moved or renamed
3. Superseded patterns: Workarounds that are no longer needed after fixes
4. Contradictions: Two entries that say different things about the same topic

Use Grep to spot-check key claims:

# Example: verify a skill count claim
grep -c "^      - id:" skills/_registry.yml
# Example: verify a file still exists
ls path/claimed/in/memory.md

Expected: A list of entries that are stale, with the correct current values.

On failure: If you can't verify a claim (e.g., it references external state you can't check), leave it but add a (unverified) note rather than silently preserving potentially wrong information.

Step 3: Decide What to Add

For new entries, apply these filters before writing:

1. Durability: Will this be true next session? Avoid session-specific context (current task, in-progress work, temporary state).
2. Non-duplication: Does CLAUDE.md or project documentation already cover this? Don't duplicate — memory is for things NOT captured elsewhere.
3. Verified: Has this been confirmed across multiple interactions, or is it a single observation? For single observations, verify against project docs before writing.
4. Actionable: Does knowing this change behavior? "The sky is blue" isn't useful. "Exit code 5 means quoting error — use temp files" changes how you work.

Exception: If the user explicitly asks to remember something, save it immediately — no need to wait for multiple confirmations.

Expected: A filtered list of entries worth adding, each meeting durability + non-duplication + verification + actionability criteria.

On failure: If unsure whether an entry is worth keeping, err toward keeping it briefly in MEMORY.md — it's easier to prune later than to rediscover.

Step 4: Extract Oversize Topics

When a section in MEMORY.md exceeds ~10-15 lines, extract it to a dedicated topic file:

1. Create <memory-dir>/<topic-name>.md with a descriptive header
2. Move the detailed content from MEMORY.md to the topic file
3. Replace the section in MEMORY.md with a 1-2 line summary and a link:

## Topic Files
- [topic-name.md](topic-name.md) — Brief description of contents

Naming conventions for topic files:

• Use lowercase kebab-case: viz-architecture.md, not VizArchitecture.md
• Name by topic, not chronology: patterns.md, not session-2024-12.md
• Group related items: combine "R debugging" and "WSL quirks" into patterns.md rather than creating one file per fact

Expected: MEMORY.md stays under 200 lines. Each topic file is self-contained and readable without MEMORY.md context.

On failure: If a topic file would be fewer than 5 lines, it's probably not worth extracting — leave it inline in MEMORY.md.

Step 5: Update MEMORY.md

Apply all changes: remove stale entries, add new entries, update counts, and ensure the Topic Files section lists all dedicated files.

MEMORY.md structure should follow this pattern:

# Project Memory

## Section 1 — High-level context
- Bullet points, concise

## Section 2 — Another topic
- Key facts only

## Topic Files
- [file.md](file.md) — What it covers

Guidelines:

• Keep each bullet to 1-2 lines maximum
• Use inline formatting (code, bold) for scanability
• Put the most frequently needed context first
• The Topic Files section should always be last

Expected: MEMORY.md is under 200 lines, accurate, and has working links to all topic files.

On failure: If you can't get under 200 lines after extraction, identify the least-frequently-used section and extract it. Every section is a candidate — even the project structure overview can go to a topic file if needed, leaving just a 1-line summary.

Step 6: Verify Integrity

Run a final check:

1. Line count: Confirm MEMORY.md is under 200 lines
2. Links: Verify every topic file referenced in MEMORY.md exists
3. Orphans: Check for topic files not referenced in MEMORY.md
4. Accuracy: Spot-check 2-3 factual claims against project state

wc -l <memory-dir>/MEMORY.md
# Check for broken links
for f in $(grep -oP '\[.*?\]\(\K[^)]+' <memory-dir>/MEMORY.md); do
  ls <memory-dir>/$f 2>/dev/null || echo "BROKEN: $f"
done
# Check for orphan files
ls <memory-dir>/*.md | grep -v MEMORY.md

Expected: Line count under 200, no broken links, no orphan files, spot-checked claims are accurate.

On failure: Fix broken links (update or remove). For orphan files, either add a reference in MEMORY.md or delete them if they're no longer relevant.

Validation

• MEMORY.md is under 200 lines
• All topic files referenced in MEMORY.md exist on disk
• No orphan .md files in memory directory (every file is linked from MEMORY.md)
• No stale counts or renamed paths in any memory file
• New entries meet the durability/non-duplication/verified/actionable criteria
• Topic files have descriptive headers and are self-contained
• MEMORY.md reads as a useful quick-reference, not a changelog

Common Pitfalls

• Memory file pollution: Writing every session observation to memory. Most findings are session-specific and don't need persisting. Apply the four filters (Step 3) before writing.
• Stale counts: Updating code but not memory. Counts (skills, agents, domains, files) drift silently. Always verify counts against the source of truth before trusting memory.
• Chronological organization: Organizing by "when I learned it" instead of "what it's about." Topic-based organization (patterns.md, viz-architecture.md) is far more useful for retrieval than date-based files.
• Duplicating CLAUDE.md: CLAUDE.md is the authoritative project instruction file. Memory should capture things NOT in CLAUDE.md — debugging insights, architecture decisions, workflow preferences, cross-project patterns.
• Over-extraction: Creating a topic file for every 3-line section. Only extract when a section exceeds ~10-15 lines. Small sections work fine inline.
• Forgetting the 200-line limit: MEMORY.md is loaded into every system prompt. Lines after 200 are silently truncated. If the file grows past this, the bottom content is effectively invisible.

Related Skills

• write-claude-md — CLAUDE.md captures project instructions; memory captures cross-session learning
• prune-agent-memory — the inverse of manage-memory: auditing, classifying, and selectively forgetting stored memories
• write-continue-here — write a structured continuation file for session handoff; complements memory as a short-term context bridge
• read-continue-here — read and act on a continuation file at session start; the consumption side of the handoff
• create-skill — new skills may produce memory-worthy patterns
• heal — self-healing may update memory as part of integration step
• meditate — meditation sessions may surface insights worth persisting