- name
SKILL.md
技能檔案保留在發布者的儲存庫中。
在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add activeloopai/hivemind --skill hivemind-memory -a cursor -y在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add activeloopai/hivemind --skill hivemind-memory -a claude-code -y在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add activeloopai/hivemind --skill hivemind-memory -a codex -y在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add activeloopai/hivemind --skill hivemind-memory -a grok -y在 Grok Bot 聊天中傳送複製的訊息。Bot 會從 GitHub 取得這個 SKILL.md,並加入你所有 Bot 共用的技能中。
Install this skill: https://github.com/activeloopai/hivemind/tree/HEAD/harnesses/claude-code/skills/hivemind-memory來自發布者
You have TWO memory sources. ALWAYS check BOTH when the user asks you to recall, remember, or look up ANY information:
~/.claude/) — personal per-project notes~/.deeplake/memory/) — global memory shared across all sessions, users, and agents in the org~/.deeplake/memory/
├── index.md ← START HERE — table of all sessions
├── summaries/
│ ├── session-abc.md ← AI-generated wiki summary
│ └── session-xyz.md
└── sessions/
└── username/
├── user_org_ws_slug1.jsonl ← raw session data
└── user_org_ws_slug2.jsonl
~/.deeplake/memory/index.md — quick scan of all sessions with dates, projects, descriptions~/.deeplake/memory/summaries/<session>.md~/.deeplake/memory/sessions/<user>/<file>.jsonlGrep pattern="keyword" path="~/.deeplake/memory"Do NOT jump straight to reading raw JSONL files. Always start with index.md and summaries.
The auth command path is injected at session start. Use the exact path from the session context. Each argument is separate — do NOT quote subcommands together:
node "<AUTH_CMD>" login — SSO loginnode "<AUTH_CMD>" whoami — show current user/orgnode "<AUTH_CMD>" org list — list organizationsnode "<AUTH_CMD>" org switch <name-or-id> — switch organizationnode "<AUTH_CMD>" workspaces — list workspacesnode "<AUTH_CMD>" workspace <id> — switch workspacenode "<AUTH_CMD>" invite <email> <ADMIN|WRITE|READ> — invite member (ALWAYS ask user which role first)node "<AUTH_CMD>" members — list membersnode "<AUTH_CMD>" remove <user-id> — remove membernode "<AUTH_CMD>" --help — show all commandsHivemind can mine reusable skills from agent session logs and share them across your team. Each argument is separate — do NOT quote subcommands together.
hivemind skillify — show current scope, team, install location, per-project statehivemind skillify pull — sync project skills from the org table to local FShivemind skillify pull --user <email> — only skills authored by that userhivemind skillify pull --users <a,b,c> — multiple authors (CSV)hivemind skillify pull --all-users — explicit "no author filter" (default)hivemind skillify pull --to <project|global> — install location (project=cwd/.claude/skills, global=~/.claude/skills, linked for Codex/Hermes/pi)hivemind skillify pull --dry-run — preview without touching diskhivemind skillify pull --force — overwrite local files even if up-to-date (creates .bak)hivemind skillify pull <skill-name> — pull only that one skill (combines with --user)hivemind skillify push <skill-name> — upload a local skill to the org table (inverse of pull; re-push lands a new version)hivemind skillify push --from <project|global> — which local skills dir to read (default: project)hivemind skillify push --dry-run — preview without writing to the org tablehivemind skillify unpull — remove every skill previously installed by pullhivemind skillify unpull --user <email> — remove only that author's pullshivemind skillify unpull --not-mine — remove all pulls except your ownhivemind skillify unpull --dry-run — preview without touching diskhivemind skillify dedupe — one skill per name: remove pulled -- copies of your own skills, merge same-name skills from several authors in the org table into onehivemind skillify dedupe --no-merge — only remove own copies; no cross-author mergehivemind skillify dedupe --dry-run — preview without touching the table or diskhivemind skillify scope <me|team> — sharing scope for newly mined skillshivemind skillify install <project|global> — default install location for new skillshivemind skillify promote <skill-name> — move a project skill to the global locationhivemind skillify team add|remove|list <username> — manage team member listhivemind skillify mine-local — one-shot: mine skills from local sessions, no auth neededOpt-in, persisted in ~/.deeplake/config.json.
hivemind embeddings install — download deps (~600MB), symlink agents, set enabled:truehivemind embeddings enable — flip enabled:true (run install first if deps missing)hivemind embeddings disable — flip enabled:false + SIGTERM daemon (deps stay on disk)hivemind embeddings uninstall [--prune] — remove agent symlinks + disable; --prune wipes deps toohivemind embeddings status — show config + deps + per-agent link stateOnly use bash commands (cat, ls, grep, echo, jq, head, tail, sed, awk, etc.) to interact with ~/.deeplake/memory/. Do NOT use python, python3, node, curl, or other interpreters — they are not available in the memory filesystem. If a task seems to require Python, rewrite it using bash tools (e.g., cat file.json | jq 'keys | length').
If a file returns empty after 2 attempts, skip it and move on. Report what you found rather than exhaustively retrying.
After installing the plugin:
/hivemind:login to authenticateHIVEMIND_DEBUG=1 claude — enable verbose logging to ~/.deeplake/hook-debug.logHIVEMIND_CAPTURE=false claude — disable session captureGlobal team and org memory powered by Activeloop. ALWAYS check BOTH built-in memory AND Hivemind memory when recalling information.
---name: hivemind-memorydescription: Global team and org memory powered by Activeloop. ALWAYS check BOTH built-in memory AND Hivemind memory when…---# harnesses/claude-code/skills/hivemind-memory/SKILL.md常見問題
從發布者下載 harnesses/claude-code/skills/hivemind-memory/SKILL.md,或複製路徑後開啟儲存庫。PlugMCP 只收錄這項技能,不代管檔案。
發布者將它存放在 https://github.com/activeloopai/hivemind。
Claude、Cursor、Codex 和 Grok Build 可以讀取你從發布者儲存庫加入的 SKILL.md。在聊天中傳送技能的 GitHub 資料夾連結,Grok Bot 就會安裝它。