@deeplake/hivemind 0.7.46 → 0.7.48

package/.claude-plugin/marketplace.json

@@ -6,18 +6,18 @@
   },
   "metadata": {
     "description": "Cloud-backed persistent shared memory for AI agents powered by Deeplake",
-    "version": "0.7.46"
+    "version": "0.7.48"
   },
   "plugins": [
     {
       "name": "hivemind",
       "description": "Persistent shared memory powered by Deeplake — captures all session activity and provides cross-session, cross-agent memory search",
-      "version": "0.7.46",
+      "version": "0.7.48",
       "source": {
         "source": "git-subdir",
         "url": "https://github.com/activeloopai/hivemind.git",
         "path": "claude-code",
-        "sha": "3ae71959dab7034fc6e8eac45d10e8ed1d4c13e5"
+        "sha": "304d5e21d39301b1941d167abdfac1b68401f15a"
       },
       "homepage": "https://github.com/activeloopai/hivemind"
     }

package/.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "hivemind",
   "description": "Cloud-backed persistent memory powered by Deeplake — read, write, and share memory across Claude Code sessions and agents",
-  "version": "0.7.46",
+  "version": "0.7.48",
   "author": {
     "name": "Activeloop",
     "url": "https://deeplake.ai"

package/README.md

@@ -285,6 +285,70 @@ hivemind skillify unpull                     # remove pulled skills
 
 Triggers, generation flow, full `pull` / `unpull` semantics, gate-CLI table per agent, env vars, logs: **[docs/SKILLIFY.md](docs/SKILLIFY.md)**.
 
+## Rules + tasks (cross-agent KPIs)
+
+Hivemind **shares team rules and tasks across every agent in the org**, injected at SessionStart so every claude-code / cursor / hermes session starts knowing them. Tasks come with LLM-generated KPIs; the agent records progress via a CLI; `tasks report` shows current/target.
+
+```bash
+# Rules — team-wide principles ("never DROP TABLE on prod")
+hivemind rules add "no DROP TABLE on prod creds"
+hivemind rules list                          # latest 10 active
+hivemind rules edit <rule-id> "<new text>"   # bumps version
+hivemind rules done <rule-id>                # mark closed
+
+# Tasks — personal or team work items, with agent-generated KPIs
+hivemind tasks add "ship the search bar" [--scope me|team] [--assign <user>]
+hivemind tasks list [--mine|--team|--all] [--status active|done|all]
+hivemind tasks edit <task-id> "<new text>"
+hivemind tasks done <task-id>
+hivemind tasks assign <task-id> <user>
+hivemind tasks progress <task-id> <kpi-id> --value N [--note "..."]
+hivemind tasks report [<task-id>]            # SUM events per KPI
+
+# Cross-agent diagnostic / pi/openclaw fallback
+hivemind context                             # print the injection block on demand
+```
+
+**What's injected at SessionStart** (claude-code, cursor, hermes — codex
+deliberately excluded to keep its user-visible TUI clean; pi/openclaw
+fall back to `hivemind context`):
+
+```text
+=== HIVEMIND RULES (N active) ===
+- <rule_id>: <text>
+(X more — run 'hivemind rules list' to see all)
+
+=== HIVEMIND TASKS (N active) ===
+[team] <task_id>: <text> ★YOU | PRs merged: 3/5 count
+[me]   <task_id>: <text>      | Lines reviewed: 75/200 lines
+(X more — run 'hivemind tasks list' to see all)
+
+=== HIVEMIND HOW-TO ===
+- Rules above are team principles. Treat any action that would violate one as a critical error and surface it to the user before proceeding.
+- Tasks above are your current work. Use 'hivemind tasks progress <task-id> <kpi-id> --value N' to record progress on a KPI.
+- Run 'hivemind rules list' / 'hivemind tasks list' for the full inventories beyond what's shown here.
+```
+
+**Auto-extract** — `gh pr merge` from inside an agent's Bash tool
+appends an event to `hivemind_task_events` automatically (failed
+merges, `--auto` enablement, interrupted commands all skipped). The
+event is "orphan" in v1 (no task binding yet); v1.1 ships
+`hivemind events attribute` for retroactive binding.
+
+**KPI generation** — `hivemind tasks add` calls Claude Sonnet to
+produce 1-3 KPIs from the task text. Failure modes (no API key,
+timeout, parse error) degrade to empty kpis; the task still INSERTs
+and the user can record progress manually.
+
+**Env vars:**
+- `HIVEMIND_RULES_TABLE`, `HIVEMIND_TASKS_TABLE`, `HIVEMIND_TASK_EVENTS_TABLE` — table names (default `hivemind_rules` / `hivemind_tasks` / `hivemind_task_events`).
+- `HIVEMIND_KPI_MODEL` — model id for KPI generation (default `claude-sonnet-4-6`).
+- `HIVEMIND_KPI_LLM=disable` — opt out of LLM KPI gen.
+- `ANTHROPIC_API_KEY` — required for KPI gen; absence is a silent no-op (kpis = `[]`).
+- `HIVEMIND_CAPTURE=false` — full read-only mode. Skips placeholder + ensure DDL; renderer still injects.
+
+Full data model, identity contract, per-agent injection status, auto-extract pipeline, KPI LLM details, and known v1 limitations: **[docs/RULES_TASKS_KPIS.md](docs/RULES_TASKS_KPIS.md)**.
+
 ## Architecture
 
 Per-agent integration mechanisms (marketplace plugin, hooks, skills, native extension) and monorepo structure: **[docs/ARCHITECTURE.md](docs/ARCHITECTURE.md)**.