docs: reposition readme front-matter

Author: SGevorg

CHANGELOG.md

@@ -1,5 +1,8 @@
 # Changelog
 
+## 1.28.11
+- Refresh README front-matter messaging. New tagline ("Shared operational brain for teams of AI coworkers") replaces the prior "self-improving context engine for teams of agents" framing. New lead emphasises a knowledge graph of the work your team has actually done, sourced from the systems where the work lives, rather than from chat history. New §1 "Your operation isn't in the chat" makes the chat-vs-records distinction explicit. New §2 closing line describes PearScarf as a multi-agent, self-evolving system with conversational and MCP interfaces. No code or behaviour changes.
+
 ## 1.28.10
 - Fix `save_issue` (linearscarf 0.1.8) to re-fetch the issue from Linear by identifier instead of relying on the payload the agent passes. `LinearGetIssueTool` truncates the issue description to 1000 chars for display; agents that read an issue then save it were passing the truncated text through, which dropped the `## For agents` YAML block on issues whose bodies exceed the cap (e.g. PEA-147 lost the YAML and fell back to Path 2 prose extraction, ending up with a generic "Shipped in pearscarf 1.28.0" fact instead of the title-anchored declared one). `SaveIssueTool`'s schema now takes only `identifier`; `execute` calls `connect.get_issue(identifier)` to fetch the full current state and feeds that to `ingest_record`. The save path becomes structurally equivalent to the polling ingester — extraction quality no longer depends on the agent's context fidelity.
 

README.md

@@ -5,7 +5,7 @@
 <h1 align="center">PearScarf</h1>
 
 <p align="center">
-  Self-improving context engine for teams of agents.
+  Shared operational brain for teams of AI coworkers.
 </p>
 
 <p align="center">
@@ -14,23 +14,23 @@
 
 ---
 
-PearScarf is a self-improving, reality-aligned context engine for teams of agents, built on a bi-temporal knowledge graph.
+PearScarf is the shared operational brain for your AI coworkers — a knowledge graph of the work your team has actually done, sourced from the systems where the work lives, not from what someone said about it in a thread. Every coworker reads from the same view; none of them rebuild context on every run.
 
-It watches your data sources — Gmail, Linear, GitHub, and more — extracts what matters, and makes it queryable by any agent in your system. One structured call. Sourced, dated, current context. No raw record processing on every run.
+It separates **observed reality** (what shipped, what's true now) from **stated intention** (commitments, plans, goals) — so coworkers don't confuse a promise with a delivery.
 
-## Why current approaches fall short
+## Your operation isn't in the chat
 
-Teams of agents deal with heterogeneous data — emails, issues, PRs, calendar, CRM — all connected in your head but siloed for your agents. Vector storage-based RAG loses those connections and when things happened. Stuffing raw records into context grows with every new source, and still leaves every agent rebuilding the same connections on every run.
+Most agent-memory tools learn from conversations — a noisy log of one assistant's exchanges with one user. Your operation lives elsewhere: in Linear issues, Gmail threads, GitHub PRs, calendar events, CRM updates, Spreadsheets. PearScarf reads from those systems directly and keeps the connections between them in a graph. Vector retrieval over raw records loses those connections; hand-maintained markdown systems go stale.
 
-## One call. Everything your agent needs to know.
+## One call. Everything your coworker needs to know.
 
-PearScarf sits between your data sources and your agents. It observes records as they arrive, extracts structured facts with full provenance, and maintains those facts over time — nothing silently overwritten, history always preserved. When an agent needs context, it asks PearScarf. One call returns everything known about an entity: current state, recent activity, open commitments, blockers.
+PearScarf watches your operational systems — Linear, Gmail, GitHub — and extracts structured facts with full provenance. Nothing silently overwritten, history always preserved. When a coworker needs context, it asks PearScarf. One call returns everything known about an entity: current state, recent activity, open commitments, blockers — sourced and dated.
 
-PearScarf is itself an agent — built and run as one. Its captures and maintaines the state of the world so your other agents don't have to.
+PearScarf is itself multi-agent, self-evolving system with conversational, mcp interfaces. It captures and maintains the state of your operation so your other coworkers don't have to.
 
 > **For MCP-compatible frameworks**
 >
-> Connect once via MCP. Every agent in your system gets access to the same shared, up-to-date context — without writing any retrieval logic.
+> Connect once via MCP. Every coworker in your fleet reads from the same shared brain.
 
 ## Expert plugin architecture
 

pearscarf/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "1.28.10"
+__version__ = "1.28.11"
 
 # Make installed expert packages importable from the experts/ folder.
 # Temporary — replaced by the registry-driven discovery in a follow-up.