Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add sideshowroberto/vfx-agent-toolkit --skill unreal-pcg-automationgit clone --depth 1 https://github.com/sideshowroberto/vfx-agent-toolkitWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/sideshowroberto/vfx-agent-toolkit/unreal-pcg-automation)<a href="https://agentmods.dev/skills/sideshowroberto/vfx-agent-toolkit/unreal-pcg-automation"><img src="https://agentmods.dev/badge/skills/sideshowroberto/vfx-agent-toolkit/unreal-pcg-automation/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/sideshowroberto/vfx-agent-toolkit/unreal-pcg-automation"><img src="https://agentmods.dev/badge/skills/sideshowroberto/vfx-agent-toolkit/unreal-pcg-automation.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00058 | $0.02586 |
| Opus 5.5 | $0.00023 | $0.01034 |
| Sonnet 5.5 | $0.00012 | $0.00517 |
| Haiku 4.5 | $0.00006 | $0.00259 |
Grade A, and why
unreal-pcg-automation scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 282 lines — stays where its author put it; the contents beside it link to each section on GitHub.
unreal-pcg-automation
Skill Version: 2.0.2 Last Updated: 2026-09-08 Dependencies: Unreal Engine 5.8+, UE 5.8 native MCP (HTTP, port 8000) Status: Production-ready
Preflight first (2026-09-08): before the first build step of any approved plan, run the
unreal-project-preflightskill (ue_preflight.py --live --features <list> --plan-block) so plugins and project settings this skill depends on are enabled in ONE restart. Mid-task "class not found / plugin not enabled" errors go back to that Step 0, never to a runtime toggle.
Execution Model: Direct MCP Connection
CRITICAL: This skill assumes Claude Code is connected to a running Unreal Engine instance via native MCP.
What This Means:
- Execute Python directly in Unreal using
mcp__ue58-mcp__execute_python_code - NO script generation - execute immediately
- Check Unreal Output Log for results (see Debugging section)
- Use multi-phased execution for complex graphs (see below)
Multi-Phased Workflow Pattern:
For complex PCG graphs, use 3 separate executions to avoid crashes and timeouts:
# PHASE 1: Create all nodes (separate execution)
graph = unreal.AssetToolsHelpers.get_asset_tools().create_asset(...)
node1, _ = graph.add_node_of_type(unreal.PCGGetLandscapeSettings)
node2, _ = graph.add_node_of_type(unreal.PCGSurfaceSamplerSettings)
# ... create all nodes
print(f"Created {len(graph.nodes)} nodes")
# PHASE 2: Connect all nodes (separate execution)
graph = unreal.load_asset('/Game/PCG/MyGraph')
graph.add_edge(graph.nodes[0], unreal.Name("Out"), graph.nodes[1], unreal.Name("Surface"))
# ... all connections (NO CODE AFTER LAST add_edge!)
# PHASE 3: Configure properties (separate execution)
graph = unreal.load_asset('/Game/PCG/MyGraph')
s1 = graph.nodes[1].get_settings() # Use .get_settings() for existing graphs
s1.points_per_squared_meter = 0.05
# ... all property configurations
print("Configuration complete")
Why 3 Phases:
- Overwrite popups can interrupt node creation
add_edge()triggers async validation (Silent Execution)- Settings access differs between creation vs existing graphs
What ships with it
15 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- REFACTORING_PLAN_v1.5.0.md 23 KB
- reference/advanced_nodes.md 12 KB
- reference/advanced_patterns.md 8.0 KB
- reference/api_limitations.md 9.3 KB
- reference/common_nodes.md 14 KB
- reference/complete_graph_template.md 7.0 KB
- reference/landscape_patches_integration.md 2.7 KB
- reference/landscape_scatter_workflow.md 12 KB
- reference/pin_discovery_patterns.md 6.5 KB
- reference/production_patterns.md 17 KB
- reference/property_verification.md 7.2 KB
- reference/silent_execution_deep_dive.md 7.9 KB
- reference/spline_workflows.md 16 KB
- reference/troubleshooting.md 9.4 KB
- reference/workflows.md 18 KB
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- today First seen · 282 lines · 58 tokens per session scan A bba3a762f12b
unreal-pcg-automation is a skill published in the GitHub repository sideshowroberto/vfx-agent-toolkit (16 stars, last pushed 22d ago), licensed MIT. It adds 58 tokens to every session and 2,586 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-11-02.
Other skills, from other repositories
ue5-debug-validation
UE5.6/UE5.7 debugging and validation workflow for logs, asset checks, and regression triage. Use when requests involve troubleshooting why gameplay does not work, validating expected output, narrowing minimal repro, and producing concrete fix steps.
scenario-unreal-gameplay
Use when building or debugging UE5 gameplay: a character with double jump or dash, abilities with cooldowns, enemy AI that patrols, chases and attacks, a health bar or HUD; choosing Blueprint vs C++, GAS or not, StateTree vs Behavior Tree, EQS or Mass; setting up Enhanced Input, Common UI or UMG; replication…
scenario-unreal-performance
Use when an Unreal Engine 5.8 project misses its frame rate or stutters: "optimize to 60 fps", "we run at 38 fps", CPU or GPU bound, game or render thread too long, hitches, traversal or shader stutter (PSO), GC spikes, camera-cut spikes, Lumen, Nanite, VSM, ray tracing, foliage, crowds or Niagara too expensive, async…
scenario-unity-gameplay
Use when building or debugging Unity gameplay systems: enemy AI that patrols, chases and attacks, NavMesh baking and agents (AI Navigation 2), feet sliding on agents, state machines or behavior trees, physics projectiles that tunnel, the physics spiral, layers and the collision matrix, CharacterController vs Rigidbody…
scenario-unity-performance
Use when a Unity 6.3 game is slow or stutters: low fps, frame spikes, hitches when entering areas or on first effects, GC spikes, too many draw calls or SetPass calls, CPU vs GPU bound questions, Profiler, Profile Analyzer, Memory Profiler, Frame Debugger or Project Auditor work, GPU Resident Drawer or SRP Batcher…
dcc-diagnostics
Infrastructure skill — DCC-agnostic observability primitives: capture screenshots, query audit logs, inspect tool performance metrics, and monitor process health. Works in any DCC environment (Maya, Blender, Houdini, Unreal, etc.) or standalone Python. Use for debugging any skill failure or verifying DCC state. Not…