メニュー
Use when the user asks to "remember project context"; manages the SEO/GEO memory lifecycle — hot-cache, active work, archive tiers, and privacy cleanup. Not for content or domain scoring — use the auditors. 项目记忆/跨会话
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/aaron-he-zhu/seo-geo-claude-skills --skill memory-managementコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
Use when the user asks to "optimize for AI citations"; improves citation readiness for ChatGPT, Perplexity, AI Overviews, Gemini, and Claude. Not for structural on-page SEO — use on-page-seo-auditor; not for net-new drafting — use seo-content-writer. AI引用优化/GEO优化/AI搜索
Use when the user asks to "optimize meta tags"; improves titles, descriptions, Open Graph, Twitter cards, and CTR test variants. Not for JSON-LD structured data — use schema-markup-generator; not for body copy — use seo-content-writer. 标题优化/元描述/CTR
Use when the user asks to "generate schema"; creates JSON-LD for FAQ, HowTo, Article, Product, and LocalBusiness rich-result candidates. Not for title/meta-description tags — use meta-tags-optimizer; not for crawl/index technical issues — use technical-seo-checker. Schema标记/结构化数据
SKILL.md を表示中
SOC 職業分類に基づく
| name | memory-management |
| description | Use when the user asks to "remember project context"; manages the SEO/GEO memory lifecycle — hot-cache, active work, archive tiers, and privacy cleanup. Not for content or domain scoring — use the auditors. 项目记忆/跨会话 |
| version | 9.9.12 |
| license | Apache-2.0 |
| compatibility | Claude Code and compatible agent-skill hosts |
| homepage | https://github.com/aaron-he-zhu/seo-geo-claude-skills |
| when_to_use | Use when reviewing, archiving, or cleaning up campaign memory. Also when the user asks to check saved findings, manage hot cache, or archive old data. |
| argument-hint | [review|archive|cleanup] |
| metadata | {"author":"aaron-he-zhu","version":"9.9.12","geo-relevance":"low"} |
This skill implements a three-tier memory system (HOT/WARM/COLD) for SEO and GEO projects. HOT memory (80 lines max) loads automatically every session via the SessionStart hook. WARM memory loads on demand per skill. COLD memory is archived data queried only when explicitly requested. The skill manages the full lifecycle: capture, promote, demote, and archive.
Manages a three-tier memory lifecycle (HOT/WARM/COLD) with automatic promotion, demotion, and archival. Also maintains open-loop tracking and cross-skill aggregation.
Start with one of these prompts. Finish with a hot-cache update plan and a handoff summary using the repository format in Skill Contract.
Set up SEO memory for [project name]
Initialize memory structure for a new [industry] website optimization project
Update memory after ranking check for [keyword group]
Refresh hot cache with latest competitor analysis findings
What are our hero keywords?
Show me the last ranking update date for [keyword category]
Look up our primary competitors and their domain authority
Promote [keyword] to hot cache
Archive stale data that hasn't been referenced in 30+ days
Add [term] to project glossary: [definition]
What does [internal jargon] mean in this project?
Expected output: a memory update plan, hot-cache changes, and a short handoff summary.
memory/hot-cache.md, memory/open-loops.md, memory/decisions.md, and related memory/ folders. Manages WARM-to-COLD archival in memory/archive/. Auditor handoff archiving (v7.1.0+): when triggered by a direct user request or an auditor's explicit "Save these results?" yes-response, append a structured block to . The Stop hook never initiates memory writes. See for the exact archive block format and rules.memory/audits/YYYY-MM.mdlast_updated date. Reference-frequency counters are NOT tracked by any hook, so the lifecycle never depends on them — see State Model.memory/hot-cache.md is within the 80-line / 25KB limit, and the affected memory paths are reported back to the user.Next Best Skill below when the project memory baseline is ready for active work.Emit the standard shape from skill-contract.md §Handoff Summary Format.
See Promotion & Demotion Rules for the full promotion/demotion table and action procedures.
This skill's behavior is reinforced by the library's claude-hook.sh hooks. What the hook actually does (do not document behavior it lacks):
memory/hot-cache.md, and when memory/open-loops.md has tracked items, appends a one-line pointer to review them for staleness. It does not compute dates or a "Quick Status" — surfacing which open loops are stale is the agent's job once pointed at the file.memory/hot-cache.md exceeds 80 lines / 25KB; enforces the auditor artifact-gate on memory/audits/*.md writes; offers an optional quality check after user-facing content edits.With tools: auto-populate from ~~SEO tool, ~~analytics, ~~search console. Without tools: ask user for keywords, competitors, metrics, campaigns, and terminology. See CONNECTORS.md.
Stop and ask the user when:
memory/decisions.md entry needed to answer a query has approved_by: skill_inferred or a missing field — surface it as ADVISORY and confirm before treating it as authoritative.Continue silently (never stop for):
When a user requests SEO memory management:
For new projects, create the directory structure defined in the State Model. Key directories: memory/ (decisions, open-loops, glossary, entities, research, content, audits, monitoring).
Templates: Hot Cache Template · Glossary Template
When a user references something unclear, follow this lookup sequence:
Step 1: Check memory/hot-cache.md (hot cache)
Step 2: Check memory/glossary.md
Step 3: Check Cold Storage
memory/archive/ first for dated YYYY-MM-DD- archived files.memory/research/, memory/audits/, or memory/monitoring/.Step 4: Ask User
If not found in any layer, ask for clarification
Log the new term in glossary if it's project-specific
Decision provenance (v8.0.1+): when loading memory/decisions.md, verify each entry has approved_by: user. Entries with approved_by: skill_inferred or missing field are treated as ADVISORY — surface to user before using as authoritative. Auditor-class skills (content-quality-auditor, domain-authority-auditor) MUST ignore non-user-approved decisions when determining verdict. See skill-contract.md §Promotion Rules.
Example lookup: User asks "Update rankings for our hero KWs" → Step 1 finds "Hero Keywords (Priority 1)" in hot-cache → extract the keyword list → run the ranking check → update memory/hot-cache.md and memory/monitoring/rank-history/YYYY-MM-DD-ranks.csv.
Reference: See Promotion & Demotion Rules for detailed promotion/demotion triggers (keywords, competitors, metrics, campaigns) and the action procedures for each.
Reference: See Update Triggers & Integration for the complete update procedures after ranking checks, competitor analyses, audits, and reports; monthly/quarterly archive routines; and integration points with all 8 connected skills (keyword-research, rank-tracker, competitor-analysis, content-gap-analysis, seo-content-writer, content-quality-auditor, domain-authority-auditor).
When invoked for review or cleanup:
memory/hot-cache.md. If >80, list oldest entries for archival.last_updated date (or file mtime) is older than 30 days; recommend archival for files older than 90 days. Age is computable from disk — reference-frequency is not tracked, so never gate on "unreferenced".name, description, and type in their frontmatter. Report any missing fields.Ask "Save these results for future sessions?" — if yes, write YYYY-MM-DD-<topic>.md to memory/. Add veto issues to memory/hot-cache.md only from auditor handoff or explicit user approval.
memory/ may store third-party personal data — entity names, founder bios, LinkedIn profiles, author/journalist names surfaced by entity-optimizer or research skills. Under GDPR Art 4(1) (applies to processing of personal data of EU/EEA/UK residents regardless of where the controller is located), these qualify as "personal data". The user is the data controller. Non-EU users without EU/EEA/UK data subjects may still face analogous obligations under CCPA/CPRA (California), PIPEDA (Canada), LGPD (Brazil), or other national regimes. Not legal advice.
memory/archive/ after 90 days unreferenced (default lifecycle)Invoke: memory-management purge <entity-name-or-slug>
This skill then:
memory/ (including memory/archive/) for the entity name, slug, or domain — grep -rF "<entity-name>" memory/ — and presents matches for confirmation.memory/hot-cache.md, WARM notes, COLD/archive files, memory/entities/<slug>.md, memory/entities/candidates.md, audit aggregates, and open loops.memory/audits/gdpr-purges.md per the GDPR Purge Log Template — required fields date, redacted_label, legal_basis, action, scope, working_tree_only: true — so there is a human-readable record of the request.Honest limitation — this edits the working tree only. If
memory/is under version control (it usually is, in the user's project repo), the subject still exists in git history. Verify withgit log -S"<entity-name>" -- memory/; true erasure from history requiresgit filter-repo/git filter-branchand is the user's responsibility — it is out of this skill's scope. Do not represent a working-tree redaction as a complete, audit-grade erasure. There is no salted-fingerprint or reingest-blocking mechanism: nothing in the hooks consults a tombstone before writing, so any such claim would be false.
Before writing a third-party person to memory/entities/, the user must have one lawful basis per GDPR Art 6 (where GDPR applies — see scope note above): consent, legitimate_interest, contract, or equivalent. Advisory — this skill does not enforce, and does not substitute for legal review.
Primary: keyword-research — seed or refresh campaign strategy with current demand signals.