sync: Merge monorepo updates (glm46, serialization tests, workflows) (#3)

## Summary
- GLM-4.6 module cleanup (dead tests removed, clippy fixes)
- 36 comprehensive serialization tests added
- Workspace documentation and brand asset updates
- CI workflow sync (SHA pinning, SIGILL fixes already in place)

**Note**: This syncs the monorepo changes with the OSS repo. The
histories diverged due to independent CI commits on both sides, but
content is equivalent.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: ReasonKit Bot <[email protected]>
Co-authored-by: Claude Opus 4.5 <[email protected]>

Author: 3 people

.claude/memories/project_memory.json

@@ -0,0 +1,56 @@
+{
+  "memories": [],
+  "manual_memories": [],
+  "realtime_memories": [
+    {
+      "type": "claude_response",
+      "content": "**Fixed.** All active MegaLogic references renamed to LaserLogic:\n\n| File | Status |\n|------|--------|\n| `~/.claude/commands/rk-logic.md` | ✅ Fixed |\n| `~/.claude/commands/rk-help.md` | ✅ Fixed |\n| `~",
+      "added_at": "2026-01-03T00:45:09.992082",
+      "source": "realtime_capture"
+    },
+    {
+      "type": "message",
+      "content": "Are these Claude Code (`claude`) commands a default part of the installation right now?",
+      "added_at": "2026-01-03T00:45:09.992160",
+      "source": "realtime_capture"
+    },
+    {
+      "type": "claude_response",
+      "content": "No. The `/rk-*` commands are **custom commands you created** - they're stored in your personal `~/.claude/commands/` directory.\n\n**Default Claude Code commands** (built-in):\n- `/help` - Show help\n- `/",
+      "added_at": "2026-01-03T00:47:38.305560",
+      "source": "realtime_capture"
+    },
+    {
+      "type": "message",
+      "content": "Yes, but are they installed as soon as someone would install ReasonKit right now? How is this for via `cargo` and how for via the one-liner install (curl + bash)?",
+      "added_at": "2026-01-03T00:47:38.305614",
+      "source": "realtime_capture"
+    },
+    {
+      "type": "claude_response",
+      "content": "Looking at the install script (`reasonkit-site/install.sh`):\n\n**Current state:**\n\n| Method | Claude Commands Installed? |\n|--------|---------------------------|\n| `cargo install reasonkit` | ❌ None |\n",
+      "added_at": "2026-01-03T00:57:39.869252",
+      "source": "realtime_capture"
+    },
+    {
+      "type": "message",
+      "content": "curl -fsSL https://reasonkit.sh/install | bash - Let me check if this script exists and what it does: --> It should be the same as the website: 'curl -fsSL https://get.reasonkit.sh | bash' --- - Is it",
+      "added_at": "2026-01-03T00:57:39.869315",
+      "source": "realtime_capture"
+    },
+    {
+      "type": "claude_response",
+      "content": "## Summary: ReasonKit Installation System Overhaul Complete ✅\n\n### Completed Tasks\n\n**1. Install Script v2.0.0 - Complete Overhaul** (`reasonkit-site/install.sh` & `install`)\n- Interactive by default ",
+      "added_at": "2026-01-03T01:33:48.525007",
+      "source": "realtime_capture"
+    },
+    {
+      "type": "message",
+      "content": "``` 4. UV Mandate Enforcement (CONS-010) - Fixed: docs/guides/INTEGRATION.md - pip → uv pip - Fixed: docs/reference/API_REFERENCE.md - pip → uv pip - Fixed: docs/getting-started/INSTALLATION_TROUBLESH",
+      "added_at": "2026-01-03T01:33:48.525108",
+      "source": "realtime_capture"
+    }
+  ],
+  "created_at": "2026-01-03T00:45:09.744834",
+  "updated_at": "2026-01-03T01:33:48.525111"
+}

.github/GOOD_FIRST_ISSUES.md

@@ -340,8 +340,8 @@ Make CLI help text more descriptive and include examples.
 .long_about(
     "Execute a ThinkTool protocol to analyze a query with structured reasoning.\n\n\
      Examples:\n\
-     rk-core think \"Should I take this job?\" --profile balanced\n\
-     rk-core think \"Is this email a phishing attempt?\" --profile quick"
+     rk think \"Should I take this job?\" --profile balanced\n\
+     rk think \"Is this email a phishing attempt?\" --profile quick"
 )
 ```
 
@@ -380,7 +380,7 @@ set -euo pipefail
 
 echo "Running ReasonKit analysis..."
 
-rk-core think \
+rk think \
     "Your question here" \
     --profile balanced \
     --output json > result.json

.github/workflows/README.md

@@ -75,12 +75,12 @@ cargo bench --bench retrieval_bench
 
 | Platform             | Target                      | Binary Name                        |
 | -------------------- | --------------------------- | ---------------------------------- |
-| Linux x86_64 (glibc) | `x86_64-unknown-linux-gnu`  | `rk-core-linux-x86_64.tar.gz`      |
-| Linux x86_64 (musl)  | `x86_64-unknown-linux-musl` | `rk-core-linux-x86_64-musl.tar.gz` |
-| Linux ARM64          | `aarch64-unknown-linux-gnu` | `rk-core-linux-aarch64.tar.gz`     |
-| macOS x86_64         | `x86_64-apple-darwin`       | `rk-core-macos-x86_64.tar.gz`      |
-| macOS ARM64          | `aarch64-apple-darwin`      | `rk-core-macos-aarch64.tar.gz`     |
-| Windows x86_64       | `x86_64-pc-windows-msvc`    | `rk-core-windows-x86_64.exe.zip`   |
+| Linux x86_64 (glibc) | `x86_64-unknown-linux-gnu`  | `rk-linux-x86_64.tar.gz`      |
+| Linux x86_64 (musl)  | `x86_64-unknown-linux-musl` | `rk-linux-x86_64-musl.tar.gz` |
+| Linux ARM64          | `aarch64-unknown-linux-gnu` | `rk-linux-aarch64.tar.gz`     |
+| macOS x86_64         | `x86_64-apple-darwin`       | `rk-macos-x86_64.tar.gz`      |
+| macOS ARM64          | `aarch64-apple-darwin`      | `rk-macos-aarch64.tar.gz`     |
+| Windows x86_64       | `x86_64-pc-windows-msvc`    | `rk-windows-x86_64.exe.zip`   |
 
 ### Release Process
 

API_DOCUMENTATION.md

@@ -1011,19 +1011,19 @@ ReasonKit Core includes a command-line interface:
 cargo build --release
 
 # Run commands
-./target/release/rk-core --help
+./target/release/rk --help
 
 # Ingest documents
-./target/release/rk-core ingest --path ./data/papers --recursive
+./target/release/rk ingest --path ./data/papers --recursive
 
 # Query the knowledge base
-./target/release/rk-core query "What is chain-of-thought prompting?" --top-k 10
+./target/release/rk query "What is chain-of-thought prompting?" --top-k 10
 
 # Get statistics
-./target/release/rk-core stats
+./target/release/rk stats
 
 # Serve as HTTP API
-./target/release/rk-core serve --port 8080
+./target/release/rk serve --port 8080
 ```
 
 ---

AUDIT_REPORT_2025-12-31.md

@@ -16,14 +16,14 @@
 
 2. **Fixed `opencode` Integration**:
    - Updated `reasonkit-core/scripts/rk-cli-wrapper.sh` to correctly handle `opencode` CLI.
-   - Added check for `rk-core` to use native `--profile` and `--query` arguments instead of injecting protocol text.
+   - Added check for `rk` to use native `--profile` and `--query` arguments instead of injecting protocol text.
    - Corrected `opencode` invocation to use `run` command instead of `-p` flag (matching `opencode --help`).
 
 ### Code Quality Audit
 
 - **Safety**: Found `unwrap()` usage in `mcp/server.rs` (tests) and `processing/chunking.rs` (safe regex/find).
 - **Performance**: `llm.rs` uses connection pooling. `executor.rs` supports parallel execution.
-- **Protocol Compliance**: `rk-cli-wrapper.sh` now correctly respects the `rk-core` interface.
+- **Protocol Compliance**: `rk-cli-wrapper.sh` now correctly respects the `rk` interface.
 
 ### Recommendations
 

Cargo.toml

@@ -1,8 +1,8 @@
 [package]
 name = "reasonkit-core"
-version = "0.1.0"
+version = "0.1.3"
 edition = "2021"
-rust-version = "1.74"
+rust-version = "1.75"
 authors = ["ReasonKit Team <[email protected]>"]
 description = "The Reasoning Engine — Auditable Reasoning for Production AI | Rust-Native | Turn Prompts into Protocols"
 license = "Apache-2.0"
@@ -65,7 +65,7 @@ crate-type = ["cdylib", "rlib"]
 doctest = false
 
 [[bin]]
-name = "rk-core"
+name = "rk"
 path = "src/main.rs"
 
 [[bin]]
@@ -91,6 +91,11 @@ name = "m2_integration_example"
 path = "examples/m2_integration_example.rs"
 required-features = ["minimax"]
 
+[[example]]
+name = "glm46_quick_start"
+path = "examples/glm46_quick_start.rs"
+required-features = ["glm46"]
+
 # ============================================================================
 # DEPENDENCIES
 # ============================================================================
@@ -118,11 +123,11 @@ regex = "1.10"
 qdrant-client = { version = "1.10", optional = true }
 
 # Full-text Search (optional - use reasonkit-mem for full memory features)
-tantivy = { version = "0.22", optional = true }
+tantivy = { version = "0.25", optional = true }
 
 # Memory Infrastructure (optional - provides storage, retrieval, embeddings)
 # Version required for crates.io publishing; path used for workspace development
-reasonkit-mem = { version = "0.1.0", optional = true }
+reasonkit-mem = { version = "0.1.1", optional = true }
 
 # Async Runtime
 tokio = { version = "1", features = ["full"] }
@@ -262,6 +267,11 @@ name = "harness"
 harness = false
 required-features = ["memory"]
 
+[[bench]]
+name = "glm46_benchmark"
+harness = false
+required-features = ["glm46"]
+
 # ============================================================================
 # FEATURES
 # ============================================================================
@@ -300,6 +310,10 @@ code-intelligence = []
 # Status: Incomplete - do not use in production
 minimax = []
 
+# EXPERIMENTAL: GLM-4.6 integration module
+# Status: Incomplete - do not use in production
+glm46 = []
+
 # Aesthetic Expression Mastery System - M2-enhanced UI/UX assessment
 aesthetic = ["md5"]
 

GETTING_STARTED.md

@@ -27,7 +27,7 @@ curl -fsSL https://reasonkit.sh/install | bash
 export ANTHROPIC_API_KEY="sk-ant-..."
 
 # Run
-rk-core think "Should I use microservices?" --profile quick
+rk think "Should I use microservices?" --profile quick
 ```
 
 **That's it.** You now have structured AI reasoning.