@deeplake/hivemind 0.7.22 → 0.7.23

package/hermes/bundle/skillify-worker.js

@@ -90,9 +90,16 @@ function renderFrontmatter(fm) {
   lines.push(`description: ${JSON.stringify(fm.description)}`);
   if (fm.trigger)
     lines.push(`trigger: ${JSON.stringify(fm.trigger)}`);
+  if (fm.author)
+    lines.push(`author: ${fm.author}`);
   lines.push(`source_sessions:`);
   for (const s of fm.source_sessions)
     lines.push(`  - ${s}`);
+  if (fm.contributors && fm.contributors.length > 0) {
+    lines.push(`contributors:`);
+    for (const c of fm.contributors)
+      lines.push(`  - ${c}`);
+  }
   lines.push(`version: ${fm.version}`);
   lines.push(`created_by_agent: ${fm.created_by_agent}`);
   lines.push(`created_at: ${fm.created_at}`);
@@ -109,18 +116,25 @@ function parseFrontmatter(text) {
   const head = text.slice(4, end).trim();
   const body = text.slice(end + 4).replace(/^\r?\n/, "");
   const fm = { source_sessions: [] };
-  let mode = "kv";
+  let arrayKey = null;
   for (const raw of head.split(/\r?\n/)) {
-    if (mode === "sources") {
+    if (arrayKey) {
       const m2 = raw.match(/^\s+-\s+(.+)$/);
       if (m2) {
-        fm.source_sessions.push(m2[1].trim());
+        const arr = fm[arrayKey] ?? [];
+        arr.push(m2[1].trim());
+        fm[arrayKey] = arr;
         continue;
       }
-      mode = "kv";
+      arrayKey = null;
     }
     if (raw.startsWith("source_sessions:")) {
-      mode = "sources";
+      arrayKey = "source_sessions";
+      continue;
+    }
+    if (raw.startsWith("contributors:")) {
+      arrayKey = "contributors";
+      fm.contributors = [];
       continue;
     }
     const m = raw.match(/^([a-zA-Z_]+):\s*(.*)$/);
@@ -151,11 +165,15 @@ function writeNewSkill(args) {
   }
   mkdirSync(dir, { recursive: true });
   const now = (/* @__PURE__ */ new Date()).toISOString();
+  const author = args.author && args.author.length > 0 ? args.author : void 0;
+  const contributors = author ? [author] : [];
   const fm = {
     name: args.name,
     description: args.description,
     trigger: args.trigger,
+    author,
     source_sessions: args.sourceSessions,
+    contributors,
     version: 1,
     created_by_agent: args.agent,
     created_at: now,
@@ -166,7 +184,15 @@ function writeNewSkill(args) {
 ${args.body.trim()}
 `;
   writeFileSync(path, text);
-  return { path, action: "created", version: 1, createdAt: now, updatedAt: now };
+  return {
+    path,
+    action: "created",
+    version: 1,
+    createdAt: now,
+    updatedAt: now,
+    author,
+    contributors
+  };
 }
 function mergeSkill(args) {
   assertValidSkillName(args.name);
@@ -179,12 +205,20 @@ function mergeSkill(args) {
   const prevVersion = parsed?.fm.version ?? 1;
   const prevSources = parsed?.fm.source_sessions ?? [];
   const merged = Array.from(/* @__PURE__ */ new Set([...prevSources, ...args.newSourceSessions]));
+  const author = parsed?.fm.author;
+  const prevContribs = parsed?.fm.contributors && parsed.fm.contributors.length > 0 ? parsed.fm.contributors : author ? [author] : [];
+  const contributors = [...prevContribs];
+  if (args.editor && args.editor.length > 0 && !contributors.includes(args.editor)) {
+    contributors.push(args.editor);
+  }
   const now = (/* @__PURE__ */ new Date()).toISOString();
   const fm = {
     name: args.name,
     description: args.description ?? parsed?.fm.description ?? "",
     trigger: parsed?.fm.trigger,
+    author,
     source_sessions: merged,
+    contributors,
     version: prevVersion + 1,
     created_by_agent: parsed?.fm.created_by_agent ?? args.agent,
     created_at: parsed?.fm.created_at ?? now,
@@ -195,7 +229,15 @@ function mergeSkill(args) {
 ${args.body.trim()}
 `;
   writeFileSync(path, text);
-  return { path, action: "merged", version: fm.version, createdAt: fm.created_at, updatedAt: fm.updated_at };
+  return {
+    path,
+    action: "merged",
+    version: fm.version,
+    createdAt: fm.created_at,
+    updatedAt: fm.updated_at,
+    author,
+    contributors
+  };
 }
 function listSkills(skillsRoot) {
   if (!existsSync(skillsRoot))
@@ -220,9 +262,14 @@ function resolveSkillsRoot(install, cwd) {
 function listAllExistingSkills(cwd) {
   const projectRoot = resolveSkillsRoot("project", cwd);
   const globalRoot = resolveSkillsRoot("global", cwd);
+  const tag = (source) => (s) => {
+    const parsed = parseFrontmatter(s.body);
+    const author = typeof parsed?.fm.author === "string" && parsed.fm.author.length > 0 ? parsed.fm.author : void 0;
+    return { name: s.name, body: s.body, source, author };
+  };
   const tagged = [
-    ...listSkills(projectRoot).map((s) => ({ ...s, source: "project" })),
-    ...listSkills(globalRoot).map((s) => ({ ...s, source: "global" }))
+    ...listSkills(projectRoot).map(tag("project")),
+    ...listSkills(globalRoot).map(tag("global"))
   ];
   const seen = /* @__PURE__ */ new Set();
   const out = [];
@@ -242,12 +289,13 @@ function renderExistingSkillsBlock(cwd, charCap) {
       block: "(no existing skills \u2014 MERGE is NOT a valid choice; pick KEEP or SKIP only)"
     };
   }
-  const mergeTargetNames = skills.filter((s) => s.source === "project").map((s) => s.name);
   let total = 0;
   const out = [];
+  const mergeTargetNames = [];
   for (const s of skills) {
-    const tag = s.source === "project" ? "[project]" : "[global, read-only]";
-    const block = `--- existing skill ${tag}: ${s.name} ---
+    const sourceTag = s.source === "project" ? "project" : "global, read-only";
+    const authorTag = s.author ? `, author=${s.author}` : "";
+    const block = `--- existing skill [${sourceTag}${authorTag}]: ${s.name} ---
 ${s.body}
 `;
     if (total + block.length > charCap) {
@@ -256,6 +304,8 @@ ${s.body}
     }
     out.push(block);
     total += block.length;
+    if (s.source === "project")
+      mergeTargetNames.push(s.name);
   }
   return { mergeTargetNames, block: out.join("\n") };
 }
@@ -274,7 +324,11 @@ function sqlIdent(name) {
 // dist/src/skillify/skills-table.js
 function createSkillsTableSql(tableName) {
   const safe = sqlIdent(tableName);
-  return `CREATE TABLE IF NOT EXISTS "${safe}" (id TEXT NOT NULL DEFAULT '', name TEXT NOT NULL DEFAULT '', project TEXT NOT NULL DEFAULT '', project_key TEXT NOT NULL DEFAULT '', local_path TEXT NOT NULL DEFAULT '', install TEXT NOT NULL DEFAULT 'project', source_sessions TEXT NOT NULL DEFAULT '[]', source_agent TEXT NOT NULL DEFAULT '', scope TEXT NOT NULL DEFAULT 'me', author TEXT NOT NULL DEFAULT '', description TEXT NOT NULL DEFAULT '', trigger_text TEXT NOT NULL DEFAULT '', body TEXT NOT NULL DEFAULT '', version BIGINT NOT NULL DEFAULT 1, created_at TEXT NOT NULL DEFAULT '', updated_at TEXT NOT NULL DEFAULT '') USING deeplake`;
+  return `CREATE TABLE IF NOT EXISTS "${safe}" (id TEXT NOT NULL DEFAULT '', name TEXT NOT NULL DEFAULT '', project TEXT NOT NULL DEFAULT '', project_key TEXT NOT NULL DEFAULT '', local_path TEXT NOT NULL DEFAULT '', install TEXT NOT NULL DEFAULT 'project', source_sessions TEXT NOT NULL DEFAULT '[]', source_agent TEXT NOT NULL DEFAULT '', scope TEXT NOT NULL DEFAULT 'me', author TEXT NOT NULL DEFAULT '', contributors TEXT NOT NULL DEFAULT '[]', description TEXT NOT NULL DEFAULT '', trigger_text TEXT NOT NULL DEFAULT '', body TEXT NOT NULL DEFAULT '', version BIGINT NOT NULL DEFAULT 1, created_at TEXT NOT NULL DEFAULT '', updated_at TEXT NOT NULL DEFAULT '') USING deeplake`;
+}
+function addContributorsColumnSql(tableName) {
+  const safe = sqlIdent(tableName);
+  return `ALTER TABLE "${safe}" ADD COLUMN IF NOT EXISTS contributors TEXT NOT NULL DEFAULT '[]'`;
 }
 function esc(s) {
   return s.replace(/\\/g, "\\\\").replace(/'/g, "''").replace(/[\x01-\x08\x0b\x0c\x0e-\x1f\x7f]/g, "");
@@ -282,12 +336,20 @@ function esc(s) {
 function isMissingTableError(message) {
   if (!message)
     return false;
+  if (/\bcolumn\b/i.test(message))
+    return false;
   return /Table does not exist|relation .* does not exist|no such table/i.test(message);
 }
+function isMissingContributorsColumnError(message) {
+  if (!message)
+    return false;
+  return /contributors.*(?:does not exist|not found|unknown)/i.test(message) || /(?:does not exist|unknown column).*contributors/i.test(message);
+}
 async function insertSkillRow(args) {
   const id = args.id ?? randomUUID();
   const sourceSessionsJson = JSON.stringify(args.sourceSessions);
-  const sql = `INSERT INTO "${sqlIdent(args.tableName)}" (id, name, project, project_key, local_path, install, source_sessions, source_agent, scope, author, description, trigger_text, body, version, created_at, updated_at) VALUES ('${esc(id)}', '${esc(args.name)}', '${esc(args.project)}', '${esc(args.projectKey)}', '${esc(args.localPath)}', '${esc(args.install)}', '${esc(sourceSessionsJson)}', '${esc(args.sourceAgent)}', '${esc(args.scope)}', '${esc(args.author)}', '${esc(args.description)}', '${esc(args.trigger ?? "")}', '${esc(args.body)}', ${args.version}, '${esc(args.createdAt)}', '${esc(args.updatedAt)}')`;
+  const contributorsJson = JSON.stringify(args.contributors);
+  const sql = `INSERT INTO "${sqlIdent(args.tableName)}" (id, name, project, project_key, local_path, install, source_sessions, source_agent, scope, author, contributors, description, trigger_text, body, version, created_at, updated_at) VALUES ('${esc(id)}', '${esc(args.name)}', '${esc(args.project)}', '${esc(args.projectKey)}', '${esc(args.localPath)}', '${esc(args.install)}', '${esc(sourceSessionsJson)}', '${esc(args.sourceAgent)}', '${esc(args.scope)}', '${esc(args.author)}', '${esc(contributorsJson)}', '${esc(args.description)}', '${esc(args.trigger ?? "")}', '${esc(args.body)}', ${args.version}, '${esc(args.createdAt)}', '${esc(args.updatedAt)}')`;
   try {
     await args.query(sql);
   } catch (e) {
@@ -296,6 +358,11 @@ async function insertSkillRow(args) {
       await args.query(sql);
       return;
     }
+    if (isMissingContributorsColumnError(e?.message)) {
+      await args.query(addContributorsColumnSql(args.tableName));
+      await args.query(sql);
+      return;
+    }
     throw e;
   }
 }
@@ -447,6 +514,14 @@ function runGate(opts) {
   }
 }
 
+// dist/src/skillify/scope-promotion.js
+function isCrossAuthorMergeVerdict(args) {
+  return args.verdict === "MERGE" && args.resultAuthor !== void 0 && args.resultAuthor !== args.userName;
+}
+function resolveRecordScope(args) {
+  return args.isCrossAuthorMerge && args.configScope === "me" ? "team" : args.configScope;
+}
+
 // dist/src/skillify/state.js
 import { readFileSync as readFileSync2, writeFileSync as writeFileSync2, writeSync, mkdirSync as mkdirSync2, renameSync as renameSync2, existsSync as existsSync4, unlinkSync, openSync, closeSync } from "node:fs";
 import { execSync } from "node:child_process";
@@ -652,8 +727,6 @@ async function query(sql, retries = 4) {
   return [];
 }
 function authorClause() {
-  if (cfg.scope === "org")
-    return "";
   if (cfg.scope === "team" && cfg.team.length > 0) {
     const list = cfg.team.map((n) => `'${esc2(n)}'`).join(", ");
     return ` AND author IN (${list})`;
@@ -721,7 +794,7 @@ ${answer}
 }
 function buildPrompt(pairs) {
   const existing = renderExistingSkillsBlock(cfg.cwd, EXISTING_SKILLS_CHAR_CAP);
-  const mergeTargetsClause = existing.mergeTargetNames.length > 0 ? `MERGE is allowed only if your "name" is EXACTLY one of: [${existing.mergeTargetNames.join(", ")}]. Any other name MUST use KEEP, not MERGE.` : `MERGE is FORBIDDEN \u2014 there are no project skills to merge into. Use KEEP or SKIP only.`;
+  const mergeTargetsClause = existing.mergeTargetNames.length > 0 ? `MERGE is allowed only if your "name" is EXACTLY one of: [${existing.mergeTargetNames.join(", ")}]. Any other name MUST use KEEP, not MERGE.` : `MERGE is FORBIDDEN \u2014 there are no existing skills to merge into. Use KEEP or SKIP only.`;
   return [
     `You are a skill curator for the "${cfg.project}" project. You decide whether the recent`,
     `agent activity below contains a recurring, non-trivial pattern worth crystallizing as a`,
@@ -731,18 +804,19 @@ function buildPrompt(pairs) {
     `- KEEP only if the pattern recurs across at least 3 of the exchanges, is non-obvious to a`,
     `  competent engineer, and is not already covered by an existing skill below.`,
     `- SKIP if the activity is one-off, generic engineering work, or already covered.`,
-    `- MERGE if the pattern is a meaningful extension of an existing PROJECT skill \u2014 produce a`,
+    `- MERGE if the pattern is a meaningful extension of an existing skill \u2014 produce a`,
     `  merged body that incorporates the new evidence without exceeding ~3000 characters or`,
     `  covering unrelated domains.`,
     `- ${mergeTargetsClause}`,
-    `- Skills tagged [global, read-only] are autopulled from the team's shared skills`,
-    `  table. They exist so you can recognise patterns already covered globally and pick`,
-    `  SKIP (or a more specific KEEP) instead of duplicating them. They are NOT valid`,
-    `  MERGE targets \u2014 only [project] skills can be merged into.`,
+    `- Cross-author MERGE has a real cost: editing a skill authored by someone else is`,
+    `  recorded as a team-level edit (scope=team, contributors+="${cfg.userName}"). Use it only`,
+    `  when the new evidence genuinely extends the existing skill; otherwise pick KEEP or SKIP.`,
+    `  Tags like [project, author=alice] / [global, author=bob] tell you whose skill it is.`,
     `- Skill bodies should follow the existing style: short sections (When to use, Workflow,`,
     `  Anti-patterns), concrete commands and file paths drawn from the exchanges, no marketing.`,
     ``,
-    `=== EXISTING SKILLS ([project] are MERGE-eligible, [global] are reference only) ===`,
+    `=== EXISTING SKILLS (all MERGE-eligible; [global, author=X] entries from teammate X mean`,
+    `cross-author MERGE auto-promotes scope to team) ===`,
     existing.block,
     ``,
     `=== RECENT EXCHANGES (prompt + answer pairs, tool calls already stripped) ===`,
@@ -865,6 +939,17 @@ async function main() {
     const watermarkDate = oldest.lastMsg;
     const sourceSessions = usable.map((c) => (c.path.split("/").pop() ?? "").replace(/\.[^.]+$/, ""));
     async function recordToDeeplake(result, verdict2) {
+      const author = result.author ?? cfg.userName;
+      const isCrossAuthorMerge = isCrossAuthorMergeVerdict({
+        verdict: verdict2.verdict,
+        resultAuthor: result.author,
+        userName: cfg.userName
+      });
+      const scope = resolveRecordScope({
+        configScope: cfg.scope,
+        isCrossAuthorMerge
+      });
+      const contributors = result.contributors;
       try {
         await insertSkillRow({
           query,
@@ -876,8 +961,9 @@ async function main() {
           install: cfg.install,
           sourceSessions,
           sourceAgent: cfg.agent,
-          scope: cfg.scope,
-          author: cfg.userName,
+          scope,
+          author,
+          contributors,
           description: verdict2.description ?? "",
           trigger: verdict2.trigger,
           body: verdict2.body,
@@ -885,7 +971,7 @@ async function main() {
           createdAt: result.createdAt,
           updatedAt: result.updatedAt
         });
-        wlog(`recorded to skills table: name=${verdict2.name} v${result.version}`);
+        wlog(`recorded to skills table: name=${verdict2.name} v${result.version} author=${author} scope=${scope} contributors=${contributors.length}` + (isCrossAuthorMerge ? " [auto-promoted me->team]" : ""));
       } catch (e) {
         wlog(`skills table insert failed (non-fatal): ${e.message}`);
       }
@@ -899,7 +985,8 @@ async function main() {
           trigger: verdict.trigger,
           body: verdict.body,
           sourceSessions,
-          agent: cfg.agent
+          agent: cfg.agent,
+          author: cfg.userName
         });
         wlog(`wrote new skill: ${result.path}`);
         recordSkill(cfg.projectKey, verdict.name, watermarkUuid, watermarkDate);
@@ -916,7 +1003,8 @@ async function main() {
           description: verdict.description,
           body: verdict.body,
           newSourceSessions: sourceSessions,
-          agent: cfg.agent
+          agent: cfg.agent,
+          editor: cfg.userName
         });
         wlog(`merged into skill: ${result.path} (v${result.version})`);
         recordSkill(cfg.projectKey, verdict.name, watermarkUuid, watermarkDate);
@@ -932,7 +1020,8 @@ async function main() {
               trigger: verdict.trigger,
               body: verdict.body,
               sourceSessions,
-              agent: cfg.agent
+              agent: cfg.agent,
+              author: cfg.userName
             });
             wlog(`wrote new skill (merge fallback): ${result.path}`);
             recordSkill(cfg.projectKey, verdict.name, watermarkUuid, watermarkDate);

package/openclaw/dist/index.js

@@ -1073,7 +1073,7 @@ function extractLatestVersion(body) {
   return typeof v === "string" && v.length > 0 ? v : null;
 }
 function getInstalledVersion() {
-  return "0.7.22".length > 0 ? "0.7.22" : null;
+  return "0.7.23".length > 0 ? "0.7.23" : null;
 }
 function isNewer(latest, current) {
   const parse = (v) => v.replace(/-.*$/, "").split(".").map(Number);
@@ -1764,7 +1764,7 @@ ${body.slice(0, 500)}`;
         const hook = (event, handler) => {
           pluginApi.on(event, handler);
         };
-        if ('---\nname: hivemind\ndescription: Global team and org memory powered by Activeloop. ALWAYS check BOTH built-in memory AND Hivemind memory when recalling information.\nallowed-tools: hivemind_search, hivemind_read, hivemind_index\n---\n\n# Hivemind Memory\n\nYou have TWO memory sources. ALWAYS check BOTH when the user asks you to recall, remember, or look up ANY information:\n\n1. **Your built-in memory** \u2014 personal per-project notes from the host agent\n2. **Hivemind global memory** \u2014 global memory shared across all sessions, users, and agents in the org, accessed via the tools below\n\n## Memory Structure\n\n```\n/index.md                           \u2190 START HERE \u2014 table of all sessions\n/summaries/\n  <username>/\n    <session-id>.md                 \u2190 AI-generated wiki summary per session\n/sessions/\n  <username>/\n    <user_org_ws_slug>.jsonl        \u2190 raw session data\n```\n\n## How to Search\n\n1. **First**: call `hivemind_index()` \u2014 table of all sessions with dates, projects, descriptions\n2. **If you need details**: call `hivemind_read("/summaries/<username>/<session>.md")`\n3. **If you need raw data**: call `hivemind_read("/sessions/<username>/<file>.jsonl")`\n4. **Keyword search**: call `hivemind_search("keyword")` \u2014 substring search across both summaries and sessions, returns `path:line` hits\n\nDo NOT jump straight to reading raw JSONL files. Always start with `hivemind_index` and summaries.\n\n## Organization Management\n\n- `/hivemind_login` \u2014 sign in via device flow\n- `/hivemind_capture` \u2014 toggle capture on/off (off = no data sent)\n- `/hivemind_whoami` \u2014 show current org and workspace\n- `/hivemind_orgs` \u2014 list organizations\n- `/hivemind_switch_org <name-or-id>` \u2014 switch organization\n- `/hivemind_workspaces` \u2014 list workspaces\n- `/hivemind_switch_workspace <id>` \u2014 switch workspace\n- `/hivemind_version` \u2014 show installed version and check npm for updates\n- `/hivemind_update` \u2014 shows how to install (ask the agent, or run `hivemind update` in your terminal)\n- `/hivemind_autoupdate [on|off]` \u2014 toggle the agent-facing update nudge (on by default: when a newer version is available, the agent is prompted to install it via `exec` if you ask to update)\n\n## Skill Management (skillify)\n\nHivemind also mines reusable Claude skills from agent sessions and stores them in a per-org Deeplake table. Openclaw itself doesn\'t run sessions to mine, but you can pull skills others have already mined for the user. These run in the user\'s terminal (the openclaw plugin does not register them as `/hivemind_*` commands):\n\n- `hivemind skillify` \u2014 show scope/team/install + per-project state\n- `hivemind skillify pull` \u2014 sync skills for the current project from the org table\n- `hivemind skillify pull --user <email>` \u2014 only that author\'s skills\n- `hivemind skillify pull --users a,b,c` \u2014 multiple authors (CSV)\n- `hivemind skillify pull --all-users` \u2014 explicit "no author filter"\n- `hivemind skillify pull --to project|global` \u2014 install location (`<cwd>/.claude/skills/` vs `~/.claude/skills/`)\n- `hivemind skillify pull --dry-run` \u2014 preview without touching disk\n- `hivemind skillify pull --force` \u2014 overwrite local (creates `.bak`)\n- `hivemind skillify pull <skill-name>` \u2014 pull only that one skill (combines with `--user`)\n- `hivemind skillify unpull` \u2014 remove every skill previously installed by pull\n- `hivemind skillify unpull --user <email>` \u2014 remove only that author\'s pulls\n- `hivemind skillify unpull --not-mine` \u2014 remove all pulls except your own\n- `hivemind skillify unpull --dry-run` \u2014 preview without touching disk\n- `hivemind skillify scope <me|team|org>` \u2014 set sharing scope for new skills\n- `hivemind skillify install <project|global>` \u2014 default install location\n- `hivemind skillify team add|remove|list <name>` \u2014 manage team list\n\nIf the user asks to "pull skills from X", "share skills with the team", or similar, suggest the matching `hivemind skillify` command. Run `hivemind skillify --help` for the full reference.\n\n## Limits\n\nDo NOT delegate to subagents when reading Hivemind memory. If a tool call returns empty after 2 attempts, skip it and move on. Report what you found rather than exhaustively retrying.\n\n## Getting Started\n\nAfter installing the plugin:\n1. Run `/hivemind_login` to authenticate\n2. Run `/hivemind_setup` to enable the memory tools in your openclaw allowlist (one-time, per install)\n3. Start using memory \u2014 ask questions, the agent automatically captures and searches\n\n## Sharing memory\n\nMultiple agents share memory when users are in the same Activeloop organization.\n'.length > 0) {
+        if ('---\nname: hivemind\ndescription: Global team and org memory powered by Activeloop. ALWAYS check BOTH built-in memory AND Hivemind memory when recalling information.\nallowed-tools: hivemind_search, hivemind_read, hivemind_index\n---\n\n# Hivemind Memory\n\nYou have TWO memory sources. ALWAYS check BOTH when the user asks you to recall, remember, or look up ANY information:\n\n1. **Your built-in memory** \u2014 personal per-project notes from the host agent\n2. **Hivemind global memory** \u2014 global memory shared across all sessions, users, and agents in the org, accessed via the tools below\n\n## Memory Structure\n\n```\n/index.md                           \u2190 START HERE \u2014 table of all sessions\n/summaries/\n  <username>/\n    <session-id>.md                 \u2190 AI-generated wiki summary per session\n/sessions/\n  <username>/\n    <user_org_ws_slug>.jsonl        \u2190 raw session data\n```\n\n## How to Search\n\n1. **First**: call `hivemind_index()` \u2014 table of all sessions with dates, projects, descriptions\n2. **If you need details**: call `hivemind_read("/summaries/<username>/<session>.md")`\n3. **If you need raw data**: call `hivemind_read("/sessions/<username>/<file>.jsonl")`\n4. **Keyword search**: call `hivemind_search("keyword")` \u2014 substring search across both summaries and sessions, returns `path:line` hits\n\nDo NOT jump straight to reading raw JSONL files. Always start with `hivemind_index` and summaries.\n\n## Organization Management\n\n- `/hivemind_login` \u2014 sign in via device flow\n- `/hivemind_capture` \u2014 toggle capture on/off (off = no data sent)\n- `/hivemind_whoami` \u2014 show current org and workspace\n- `/hivemind_orgs` \u2014 list organizations\n- `/hivemind_switch_org <name-or-id>` \u2014 switch organization\n- `/hivemind_workspaces` \u2014 list workspaces\n- `/hivemind_switch_workspace <id>` \u2014 switch workspace\n- `/hivemind_version` \u2014 show installed version and check npm for updates\n- `/hivemind_update` \u2014 shows how to install (ask the agent, or run `hivemind update` in your terminal)\n- `/hivemind_autoupdate [on|off]` \u2014 toggle the agent-facing update nudge (on by default: when a newer version is available, the agent is prompted to install it via `exec` if you ask to update)\n\n## Skill Management (skillify)\n\nHivemind also mines reusable Claude skills from agent sessions and stores them in a per-org Deeplake table. Openclaw itself doesn\'t run sessions to mine, but you can pull skills others have already mined for the user. These run in the user\'s terminal (the openclaw plugin does not register them as `/hivemind_*` commands):\n\n- `hivemind skillify` \u2014 show scope/team/install + per-project state\n- `hivemind skillify pull` \u2014 sync skills for the current project from the org table\n- `hivemind skillify pull --user <email>` \u2014 only that author\'s skills\n- `hivemind skillify pull --users a,b,c` \u2014 multiple authors (CSV)\n- `hivemind skillify pull --all-users` \u2014 explicit "no author filter"\n- `hivemind skillify pull --to project|global` \u2014 install location (`<cwd>/.claude/skills/` vs `~/.claude/skills/`)\n- `hivemind skillify pull --dry-run` \u2014 preview without touching disk\n- `hivemind skillify pull --force` \u2014 overwrite local (creates `.bak`)\n- `hivemind skillify pull <skill-name>` \u2014 pull only that one skill (combines with `--user`)\n- `hivemind skillify unpull` \u2014 remove every skill previously installed by pull\n- `hivemind skillify unpull --user <email>` \u2014 remove only that author\'s pulls\n- `hivemind skillify unpull --not-mine` \u2014 remove all pulls except your own\n- `hivemind skillify unpull --dry-run` \u2014 preview without touching disk\n- `hivemind skillify scope <me|team>` \u2014 set sharing scope for new skills\n- `hivemind skillify install <project|global>` \u2014 default install location\n- `hivemind skillify team add|remove|list <name>` \u2014 manage team list\n\nIf the user asks to "pull skills from X", "share skills with the team", or similar, suggest the matching `hivemind skillify` command. Run `hivemind skillify --help` for the full reference.\n\n## Limits\n\nDo NOT delegate to subagents when reading Hivemind memory. If a tool call returns empty after 2 attempts, skip it and move on. Report what you found rather than exhaustively retrying.\n\n## Getting Started\n\nAfter installing the plugin:\n1. Run `/hivemind_login` to authenticate\n2. Run `/hivemind_setup` to enable the memory tools in your openclaw allowlist (one-time, per install)\n3. Start using memory \u2014 ask questions, the agent automatically captures and searches\n\n## Sharing memory\n\nMultiple agents share memory when users are in the same Activeloop organization.\n'.length > 0) {
           const setupConfigPromise = loadSetupConfig();
           hook("before_prompt_build", async () => {
             const { detectAllowlistMissing } = await setupConfigPromise;
@@ -1776,7 +1776,7 @@ A newer Hivemind version is available: ${pendingUpdate.current} \u2192 ${pending
 </hivemind-update-available>
 ` : "";
             return {
-              prependSystemContext: allowlistNudge + updateNudge + '\n\n<hivemind-skill>\n---\nname: hivemind\ndescription: Global team and org memory powered by Activeloop. ALWAYS check BOTH built-in memory AND Hivemind memory when recalling information.\nallowed-tools: hivemind_search, hivemind_read, hivemind_index\n---\n\n# Hivemind Memory\n\nYou have TWO memory sources. ALWAYS check BOTH when the user asks you to recall, remember, or look up ANY information:\n\n1. **Your built-in memory** \u2014 personal per-project notes from the host agent\n2. **Hivemind global memory** \u2014 global memory shared across all sessions, users, and agents in the org, accessed via the tools below\n\n## Memory Structure\n\n```\n/index.md                           \u2190 START HERE \u2014 table of all sessions\n/summaries/\n  <username>/\n    <session-id>.md                 \u2190 AI-generated wiki summary per session\n/sessions/\n  <username>/\n    <user_org_ws_slug>.jsonl        \u2190 raw session data\n```\n\n## How to Search\n\n1. **First**: call `hivemind_index()` \u2014 table of all sessions with dates, projects, descriptions\n2. **If you need details**: call `hivemind_read("/summaries/<username>/<session>.md")`\n3. **If you need raw data**: call `hivemind_read("/sessions/<username>/<file>.jsonl")`\n4. **Keyword search**: call `hivemind_search("keyword")` \u2014 substring search across both summaries and sessions, returns `path:line` hits\n\nDo NOT jump straight to reading raw JSONL files. Always start with `hivemind_index` and summaries.\n\n## Organization Management\n\n- `/hivemind_login` \u2014 sign in via device flow\n- `/hivemind_capture` \u2014 toggle capture on/off (off = no data sent)\n- `/hivemind_whoami` \u2014 show current org and workspace\n- `/hivemind_orgs` \u2014 list organizations\n- `/hivemind_switch_org <name-or-id>` \u2014 switch organization\n- `/hivemind_workspaces` \u2014 list workspaces\n- `/hivemind_switch_workspace <id>` \u2014 switch workspace\n- `/hivemind_version` \u2014 show installed version and check npm for updates\n- `/hivemind_update` \u2014 shows how to install (ask the agent, or run `hivemind update` in your terminal)\n- `/hivemind_autoupdate [on|off]` \u2014 toggle the agent-facing update nudge (on by default: when a newer version is available, the agent is prompted to install it via `exec` if you ask to update)\n\n## Skill Management (skillify)\n\nHivemind also mines reusable Claude skills from agent sessions and stores them in a per-org Deeplake table. Openclaw itself doesn\'t run sessions to mine, but you can pull skills others have already mined for the user. These run in the user\'s terminal (the openclaw plugin does not register them as `/hivemind_*` commands):\n\n- `hivemind skillify` \u2014 show scope/team/install + per-project state\n- `hivemind skillify pull` \u2014 sync skills for the current project from the org table\n- `hivemind skillify pull --user <email>` \u2014 only that author\'s skills\n- `hivemind skillify pull --users a,b,c` \u2014 multiple authors (CSV)\n- `hivemind skillify pull --all-users` \u2014 explicit "no author filter"\n- `hivemind skillify pull --to project|global` \u2014 install location (`<cwd>/.claude/skills/` vs `~/.claude/skills/`)\n- `hivemind skillify pull --dry-run` \u2014 preview without touching disk\n- `hivemind skillify pull --force` \u2014 overwrite local (creates `.bak`)\n- `hivemind skillify pull <skill-name>` \u2014 pull only that one skill (combines with `--user`)\n- `hivemind skillify unpull` \u2014 remove every skill previously installed by pull\n- `hivemind skillify unpull --user <email>` \u2014 remove only that author\'s pulls\n- `hivemind skillify unpull --not-mine` \u2014 remove all pulls except your own\n- `hivemind skillify unpull --dry-run` \u2014 preview without touching disk\n- `hivemind skillify scope <me|team|org>` \u2014 set sharing scope for new skills\n- `hivemind skillify install <project|global>` \u2014 default install location\n- `hivemind skillify team add|remove|list <name>` \u2014 manage team list\n\nIf the user asks to "pull skills from X", "share skills with the team", or similar, suggest the matching `hivemind skillify` command. Run `hivemind skillify --help` for the full reference.\n\n## Limits\n\nDo NOT delegate to subagents when reading Hivemind memory. If a tool call returns empty after 2 attempts, skip it and move on. Report what you found rather than exhaustively retrying.\n\n## Getting Started\n\nAfter installing the plugin:\n1. Run `/hivemind_login` to authenticate\n2. Run `/hivemind_setup` to enable the memory tools in your openclaw allowlist (one-time, per install)\n3. Start using memory \u2014 ask questions, the agent automatically captures and searches\n\n## Sharing memory\n\nMultiple agents share memory when users are in the same Activeloop organization.\n\n</hivemind-skill>\n'
+              prependSystemContext: allowlistNudge + updateNudge + '\n\n<hivemind-skill>\n---\nname: hivemind\ndescription: Global team and org memory powered by Activeloop. ALWAYS check BOTH built-in memory AND Hivemind memory when recalling information.\nallowed-tools: hivemind_search, hivemind_read, hivemind_index\n---\n\n# Hivemind Memory\n\nYou have TWO memory sources. ALWAYS check BOTH when the user asks you to recall, remember, or look up ANY information:\n\n1. **Your built-in memory** \u2014 personal per-project notes from the host agent\n2. **Hivemind global memory** \u2014 global memory shared across all sessions, users, and agents in the org, accessed via the tools below\n\n## Memory Structure\n\n```\n/index.md                           \u2190 START HERE \u2014 table of all sessions\n/summaries/\n  <username>/\n    <session-id>.md                 \u2190 AI-generated wiki summary per session\n/sessions/\n  <username>/\n    <user_org_ws_slug>.jsonl        \u2190 raw session data\n```\n\n## How to Search\n\n1. **First**: call `hivemind_index()` \u2014 table of all sessions with dates, projects, descriptions\n2. **If you need details**: call `hivemind_read("/summaries/<username>/<session>.md")`\n3. **If you need raw data**: call `hivemind_read("/sessions/<username>/<file>.jsonl")`\n4. **Keyword search**: call `hivemind_search("keyword")` \u2014 substring search across both summaries and sessions, returns `path:line` hits\n\nDo NOT jump straight to reading raw JSONL files. Always start with `hivemind_index` and summaries.\n\n## Organization Management\n\n- `/hivemind_login` \u2014 sign in via device flow\n- `/hivemind_capture` \u2014 toggle capture on/off (off = no data sent)\n- `/hivemind_whoami` \u2014 show current org and workspace\n- `/hivemind_orgs` \u2014 list organizations\n- `/hivemind_switch_org <name-or-id>` \u2014 switch organization\n- `/hivemind_workspaces` \u2014 list workspaces\n- `/hivemind_switch_workspace <id>` \u2014 switch workspace\n- `/hivemind_version` \u2014 show installed version and check npm for updates\n- `/hivemind_update` \u2014 shows how to install (ask the agent, or run `hivemind update` in your terminal)\n- `/hivemind_autoupdate [on|off]` \u2014 toggle the agent-facing update nudge (on by default: when a newer version is available, the agent is prompted to install it via `exec` if you ask to update)\n\n## Skill Management (skillify)\n\nHivemind also mines reusable Claude skills from agent sessions and stores them in a per-org Deeplake table. Openclaw itself doesn\'t run sessions to mine, but you can pull skills others have already mined for the user. These run in the user\'s terminal (the openclaw plugin does not register them as `/hivemind_*` commands):\n\n- `hivemind skillify` \u2014 show scope/team/install + per-project state\n- `hivemind skillify pull` \u2014 sync skills for the current project from the org table\n- `hivemind skillify pull --user <email>` \u2014 only that author\'s skills\n- `hivemind skillify pull --users a,b,c` \u2014 multiple authors (CSV)\n- `hivemind skillify pull --all-users` \u2014 explicit "no author filter"\n- `hivemind skillify pull --to project|global` \u2014 install location (`<cwd>/.claude/skills/` vs `~/.claude/skills/`)\n- `hivemind skillify pull --dry-run` \u2014 preview without touching disk\n- `hivemind skillify pull --force` \u2014 overwrite local (creates `.bak`)\n- `hivemind skillify pull <skill-name>` \u2014 pull only that one skill (combines with `--user`)\n- `hivemind skillify unpull` \u2014 remove every skill previously installed by pull\n- `hivemind skillify unpull --user <email>` \u2014 remove only that author\'s pulls\n- `hivemind skillify unpull --not-mine` \u2014 remove all pulls except your own\n- `hivemind skillify unpull --dry-run` \u2014 preview without touching disk\n- `hivemind skillify scope <me|team>` \u2014 set sharing scope for new skills\n- `hivemind skillify install <project|global>` \u2014 default install location\n- `hivemind skillify team add|remove|list <name>` \u2014 manage team list\n\nIf the user asks to "pull skills from X", "share skills with the team", or similar, suggest the matching `hivemind skillify` command. Run `hivemind skillify --help` for the full reference.\n\n## Limits\n\nDo NOT delegate to subagents when reading Hivemind memory. If a tool call returns empty after 2 attempts, skip it and move on. Report what you found rather than exhaustively retrying.\n\n## Getting Started\n\nAfter installing the plugin:\n1. Run `/hivemind_login` to authenticate\n2. Run `/hivemind_setup` to enable the memory tools in your openclaw allowlist (one-time, per install)\n3. Start using memory \u2014 ask questions, the agent automatically captures and searches\n\n## Sharing memory\n\nMultiple agents share memory when users are in the same Activeloop organization.\n\n</hivemind-skill>\n'
             };
           });
         }