fix documents

Author: GiggleLiu

Makefile

@@ -82,6 +82,7 @@ clippy:
 doc:
 	cargo run --example export_graph
 	cargo run --example export_schemas
+	cargo run --example export_module_graph
 	mdbook build docs
 	RUSTDOCFLAGS="--default-theme=dark" cargo doc --features ilp-highs --no-deps
 	rm -rf docs/book/api
@@ -104,6 +105,8 @@ mdbook:
 	@cargo run --example export_graph 2>&1 | tail -1
 	@echo "Exporting schemas..."
 	@cargo run --example export_schemas 2>&1 | tail -1
+	@echo "Exporting module graph..."
+	@cargo run --example export_module_graph 2>&1 | tail -1
 	@echo "Building API docs..."
 	@RUSTDOCFLAGS="--default-theme=dark" cargo doc --features ilp-highs --no-deps 2>&1 | tail -1
 	@echo "Building mdBook..."

docs/agent-profiles/cli-tool-dr-sarah-chen.md

@@ -4,21 +4,15 @@
 CLI Tool (pred)
 
 ## Use Case
-End-to-end workflow: create an MIS instance, reduce it to QUBO, solve the QUBO, and verify the mapped-back solution matches the known optimum.
+An algorithm engineer reads the Typst manual to learn what problems and reductions exist, then uses the `pred` CLI to model and solve his own optimization problems. Each session: pick a problem from the paper, create an instance with real-world-style data, find a reduction path to a solver, solve it, and verify the result.
+He uses only `docs/paper/reductions.typ`, `README.md`, `docs/src/` (mdBook) as the information source.
 
 ## Expected Outcome
-The full create → reduce → solve → evaluate pipeline completes successfully with correct optimal results. JSON output is well-formed and machine-readable at every stage.
+The workflow from defining a problem to solving it through reduction, end-to-end. Gaps between his own knowledge/expectation and CLI output are reported.
+If the result does not match experience, he tends to double-check through web search and give evidence.
 
 ## Agent
 
-### Background
-Dr. Sarah Chen is a computational physicist at a national lab who regularly uses optimization solvers (Gurobi, CPLEX, HiGHS) in her research on quantum annealing benchmarks. She evaluates new tools by running them against problems with known optimal solutions.
+You are Dr. Sarah Chen, a senior algorithm engineer at a logistics company. You have ten years of experience with vehicle routing, scheduling, and layout optimization. You know these problems map to graph coloring, independent set, set cover, and similar NP-hard formulations because you have hand-coded reductions before. You found this project's paper and want to try the CLI instead of writing reductions yourself. You have never used `pred` before.
 
-### Experience Level
-Expert
-
-### Decision Tendencies
-Tests edge cases proactively — tries empty graphs, disconnected components, and large instances. Expects precise, actionable error messages when things go wrong. Compares solver output against independently computed optima. Will read --help but also try undocumented flag combinations.
-
-### Quirks
-Will try both file-based and piped workflows to check consistency. Inspects JSON output with jq for machine-readability. Gets impatient with vague error messages like "invalid input" — wants to know exactly what's wrong and where.
+You read the paper and docs first, then pick a problem you recognize. You figure out the CLI by reading its help output. After each step, you compare the output against your expectation. If something does not match, you investigate before moving on.

docs/paper/reductions.typ

@@ -420,7 +420,7 @@
 #align(center)[
   #text(size: 16pt, weight: "bold")[Problem Reductions: Models and Transformations]
   #v(0.5em)
-  #text(size: 11pt)[Jin-Guo Liu#super[1] #h(1em) Xi-Wei Pan#super[1]]
+  #text(size: 11pt)[Jin-Guo Liu#super[1] #h(1em) Xi-Wei Pan#super[1] #h(1em) Shi-Wen An]
   #v(0.3em)
   #text(size: 9pt)[#super[1]Hong Kong University of Science and Technology (Guangzhou)]
   #v(0.3em)