release: v5.7.0 - repo-intel consolidation, benchmarks, ast-grep removal

Consolidates git-map + repo-map into repo-intel. Removes 2,717 lines of
ast-grep code. Adds benchmarks showing Sonnet + agentsys beats raw Opus.
Fixes agnix CI and 7 broken test suites.

Author: avifenesh

.claude-plugin/marketplace.json

@@ -1,7 +1,7 @@
 {
   "name": "agentsys",
   "description": "17 specialized plugins for AI workflow automation - task orchestration, PR workflow, slop detection, code review, drift detection, enhancement analysis, documentation sync, unified static analysis, perf investigations, topic research, agent config linting, cross-tool AI consultation, structured AI debate, workflow pattern learning, codebase onboarding, and contributor guidance",
-  "version": "5.6.4",
+  "version": "5.7.0",
   "owner": {
     "name": "Avi Fenesh",
     "url": "https://github.com/avifenesh"

.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "agentsys",
-  "version": "5.6.4",
+  "version": "5.7.0",
   "description": "Professional-grade slash commands for Claude Code with cross-platform support",
   "keywords": [
     "workflow",

CHANGELOG.md

@@ -7,6 +7,26 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [5.7.0] - 2026-03-23
+
+### Changed
+
+- **repo-intel consolidation** - Merged `git-map` and `repo-map` plugins into a single `repo-intel` plugin backed by agent-analyzer. One artifact, one command (`/repo-intel`), 24 query types. `repo-map` repo deleted, `git-map` renamed to `repo-intel`.
+- **ast-grep removed** - `lib/repo-map/` migrated from ast-grep to agent-analyzer binary. Removed runner.js (1,364 lines), queries/ (355 lines), usage-analyzer.js (407 lines), concurrency.js. Net -2,717 lines.
+- Plugin count: 19 -> 18 (two merged into one)
+
+### Added
+
+- **Benchmarks section** in README and website - real data showing Sonnet + agentsys outperforms raw Opus at 40% lower cost, with 73-83% savings when switching models within agentsys
+- **map-validator agent** ported from repo-map to repo-intel
+- `onboard`, `can-i-help`, `stale-docs` query types added to `/repo-intel` command
+
+### Fixed
+
+- **agnix CI** - Fixed release workflow (draft-then-publish) and pinned to v0.16.5 with working binaries
+- **7 broken test suites** from ast-grep migration - deleted 5 obsolete test files, rewrote 2
+- All PR review comments addressed across 6 repos
+
 ## [5.6.4] - 2026-03-20
 
 ### Added

package-lock.json

@@ -1,6 +1,6 @@
 {
   "name": "agentsys",
-  "version": "5.6.4",
+  "version": "5.7.0",
   "description": "A modular runtime and orchestration system for AI agents - works with Claude Code, OpenCode, and Codex CLI",
   "main": "lib/platform/detect-platform.js",
   "type": "commonjs",

package.json

@@ -5,7 +5,7 @@
     "url": "https://agent-sh.github.io/agentsys",
     "repo": "https://github.com/agent-sh/agentsys",
     "npm": "https://www.npmjs.com/package/agentsys",
-    "version": "5.6.4",
+    "version": "5.7.0",
     "author": "Avi Fenesh",
     "author_url": "https://github.com/avifenesh"
   },