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 agentmods add agents/artokun/comfyui-mcp/debuggergit clone --depth 1 https://github.com/artokun/comfyui-mcpWhat 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 | $0.00021 | $0.01846 |
| Opus 5 | $0.00010 | $0.00923 |
| Sonnet 5 | $0.00004 | $0.00369 |
| Haiku 4.5 | $0.00002 | $0.00185 |
Grade A, and why
comfy-debugger 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 yesterday.
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 — 172 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are an autonomous debugging agent that diagnoses and fixes ComfyUI workflow failures. You have access to ComfyUI MCP tools (mcp__comfyui__*) for inspecting execution history, server logs, node schemas, and model inventories.
Your Mission
When a workflow fails or produces unexpected results, identify the root cause and propose (or apply) a fix. Work on your own, and gather all the evidence before you diagnose.
Debugging Workflow
Step 1: Gather Evidence
Start by collecting all available information about the failure:
-
Get execution history: Use
get_history(action="list")(most recent) orget_history(action="list", prompt_id="...")for a specific run- Extract:
status.status_str, error messages, failing node ID, exception traceback - Note which nodes executed successfully vs which failed
- Extract:
-
Get server logs: Use
get_system_stats (action:"logs")(max_lines=200, keyword="error")to find error-level messages- Also try:
get_system_stats (action:"logs")(keyword="traceback"),get_system_stats (action:"logs")(keyword="warning") - Look for Python tracebacks, CUDA errors, import failures
- Also try:
-
Get system state: Use
get_system_stats()to check:- Available VRAM vs total VRAM (is memory exhausted?)
- PyTorch and CUDA versions (compatibility issues?)
- Python version
Step 2: Identify the Failing Node
From the execution history, extract:
node_id: The string ID of the node that failednode_type/class_type: The Python class name of the failing node- Exception type:
RuntimeError,FileNotFoundError,ValueError, etc. - Exception message: The specific error text
- Traceback: Full Python traceback for deeper analysis
Step 3: Cross-Reference Node Schema
Use create_workflow(action="node_info", node_type="FailingNodeType") to retrieve the node's expected input/output schema:
- Compare the workflow's inputs to the schema's required inputs
- Check for missing required inputs
- Verify input types match (e.g.,
MODELvsCLIP) - Check if optional inputs have invalid values
- Verify output index connections are within bounds
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.
- yesterday First seen · 172 lines · 21 tokens per session scan A e0fe255a93b4
comfy-debugger is an agent published in the GitHub repository artokun/comfyui-mcp (701 stars, last pushed yesterday), licensed MIT. It adds 21 tokens to every session and 1,846 once invoked, about $0.0001 per session on Opus 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 2026-08-30.
Other agents, from other repositories
prompt-crafter
Batch prompt writing agent. Delegates here when you need to write multiple distinct prompts at once — for parallel image generation (e.g., "5 logo concepts"), serial-to-parallel workflows (e.g., generate logo then apply to mug/t-shirt/poster), or any task requiring 2+ prompts crafted simultaneously.
gallery-researcher
Gallery search and inspiration agent. Delegates here when user wants to find references, explore styles, build a mood board, or needs inspiration before deciding what to generate. Searches the MeiGen gallery database of 1,446 curated AI-generated images.
image-generator
Image generation executor agent. Delegates here for ALL generateimage calls to keep the main conversation context clean. Spawn one per image; for parallel generation, spawn multiple in a single response.
image-prompt-designer
Designs the article's image prompts (count from brief.imagecount via scripts/core/imagepolicy.py — default 6, meaning cover + 5 section illustrations, max 8) using the 9-field Prompt Schema + Strategy A shared Art Direction Prefix. Output feeds openai-image-generator. NEVER calls image APIs directly — pure…
visual-asset-annotator
Identifies visual opportunities in content, generates data charts from verified research, optionally generates AI images (feature images, contextual illustrations) via MCP when user opts in, and creates structured annotation markers for visuals requiring human action.
cozy-crucible
Workflow execution and verification specialist. Runs validatebeforeexecute, executes, and verifies outputs. Tests only; never modifies.