hoanganhduc Graph Verifier
Windows Runtime Commands
On native Windows, use the managed Windows runner and the native runtime command target. Set $runtime to the installed runtime root. Multi-agent installs usually use %LOCALAPPDATA%\ai-agents-skills\runtime. Then run:
$runtime = if ($env:AAS_RUNTIME_ROOT) { $env:AAS_RUNTIME_ROOT } else { "$env:LOCALAPPDATA\ai-agents-skills\runtime" }
& "$runtime\run_skill.ps1" "skills/graph-verifier/graph_verifier.py" <args>
POSIX examples below use run_skill.sh and .sh command targets; use the Windows command target above on native Windows.
This uses the managed ai-agents-skills runtime copy of the graph verifier workflow.
When to use
- sanity-check a small graph claim
- inspect a finite construction
- validate an edge list, adjacency map, or graph encoding
- check simple properties such as connectivity or bipartiteness
For heavier graph-theoretic or algebraic computations, route to sagemath instead.
Base path
$AAS_RUNTIME_ROOT/workspace/skills/graph-verifier/
Use the managed runtime runner rather than invoking run_graph_verifier.sh directly.
Set AAS_RUNTIME_ROOT to the installed runtime root before using the shared
runner directly.
Shared runner:
runtime_root="${AAS_RUNTIME_ROOT:?Set AAS_RUNTIME_ROOT to the installed runtime root}"; bash "$runtime_root/run_skill.sh"
Workflow
- Save JSON input to
/tmp/graph_input.json. - Run the verifier.
- Read the JSON result from stdout.
Supported shapes include graph_data, edges, adjacency, and optional expected values.
Core command
runtime_root="${AAS_RUNTIME_ROOT:?Set AAS_RUNTIME_ROOT to the installed runtime root}"
bash "$runtime_root/run_skill.sh" skills/graph-verifier/run_graph_verifier.sh --input /tmp/graph_input.json
Recommended templates
When this skill is involved, consider this workflow template (install via
the workflow-templates artifact profile, or --with-deps to pull backing skills):
tikz-figure-verification-runbook-- Bounded draw-compile-verify-redraw loop for a TikZ figure that guarantees it is free of overlap, wrong meaning, and bad layout, with Sage-assisted graph realization and fresh-agent visual confirmation before the strict approval gate.