Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/laurigates/claude-pluginsnpx agentmods add skills/laurigates/claude-plugins/comfy-workflow-jsonWrote 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/laurigates/claude-plugins/comfy-workflow-json)<a href="https://agentmods.dev/skills/laurigates/claude-plugins/comfy-workflow-json"><img src="https://agentmods.dev/badge/skills/laurigates/claude-plugins/comfy-workflow-json/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/laurigates/claude-plugins/comfy-workflow-json"><img src="https://agentmods.dev/badge/skills/laurigates/claude-plugins/comfy-workflow-json.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00052 | $0.02581 |
| Opus 5.5 | $0.00021 | $0.01032 |
| Sonnet 5.5 | $0.00010 | $0.00516 |
| Haiku 4.5 | $0.00005 | $0.00258 |
Grade A, and why
comfy-workflow-json 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 1mo ago.
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 — 203 lines — stays where its author put it; the contents beside it link to each section on GitHub.
comfy-workflow-json
Workflow JSONs routinely exceed a Read tool's size limit (a moderately
complex workflow can be 25k+ tokens). Don't chunk-read or hand-edit large
workflow files — write a Python transform script that uses json.load /
json.dump.
When to Use This Skill
| Use this skill when... | Use instead when... |
|---|---|
| A workflow JSON is too large to hand-edit, or you're splicing/rewiring nodes programmatically | Auto-arranging node positions -> comfy-workflow-layout |
| A round-trip transform is silently dropping App Mode or canvas state | Deciding whether to use a subgraph/Blueprint/App Mode at all -> comfy-subgraphs-app-mode |
Edge-list rebuild — mutate links without hand-surgery
In-place link surgery (patching the flat links array and every node's
inputs[].link / outputs[].links by hand) is error-prone — the
silent-failure mode is a link id present in one place but not the other.
The robust alternative:
- Extract edges to plain dicts:
[{src, src_slot, dst, dst_slot, type}]fromwf['links']. - Mutate that edge list (redirect an edge's
dstto a new node, drop edges touching a deleted node, append new edges) and add/remove nodes fromwf['nodes']. - Rebuild authoritatively: clear every node's
inputs[].link→Noneandoutputs[].links→[], then re-number all edges 1..N intowf['links']and re-populate both endpoints' link refs from the edge list. Setlast_link_id/last_node_id.
dst_slot indexes the node's inputs array (widget entries included);
src_slot indexes outputs. New nodes get id = max(existing)+1. To
insert node X between src and Target.model: find the edge with
dst==Target, dst_slot==0, repoint its dst to X, then append
X→Target.model. This keeps the socket-array / links-table invariant
correct by construction.
Don't template off a live working JSON
A reference/comparison builder must clone a clean, stable template, not
a workflow the user actively edits in the UI. A file you generated earlier
can be silently reshaped between sessions (extra LoRAs, a swapped
VAE-decode, a converted negative) — basing new variants on it inherits that
drift and breaks structural assumptions a from-scratch builder made (e.g. a
ConditioningZeroOut lookup can raise if the template gained unexpected
nodes). Pick one canonical clean template, hardcode the per-variant deltas,
and never read a user's live working copy as a template.
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.
- 1mo ago First seen · 203 lines · 52 tokens per session scan A 079b67e03965
comfy-workflow-json is a skill published in the GitHub repository laurigates/claude-plugins (60 stars, last pushed yesterday), licensed MIT. It adds 52 tokens to every session and 2,581 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 2026-09-03.
Other skills, from other repositories
extract-ssot
Deduplicate repeated markdown (rule files, skill bodies, ADRs, docs) into one named source of truth and migrate every call site to cite it by exact heading. Use when the same prose, literal, or concept recurs across files: 'DRY this prose', 'extract a shared rule', 'single source of truth for X', a value-bump diff…
feature-branch-pr-writing
Use this skill when the user asks to write, draft, create, or improve a PR description for a Shopware core repository PR — AND that PR targets a non-trunk feature branch (not trunk itself). Trigger phrases like "write a PR description", "draft the PR", "what should I put in the PR body". The skill detects the target…
repo-activity
Scan all git repositories under /repos and report recent activity — last commit age, branch, uncommitted changes — in age-bucketed tables. Use when the user asks for a repo activity overview, "what have I been working on", portfolio status, or which repos are active/dormant.
oslog
Read, stream, and analyze Apple unified logs (OSLog) for iOS/macOS apps. Auto-detects subsystem from Logger usage or bundle identifier. Supports live log show, real-time streaming, and .logarchive analysis with full predicate reference. Use when the user asks to check app logs, stream console output, debug with OSLog…
renovate-triage
Use when triaging open Renovate PRs across your own repos into merge / close / defer. Activates on "renovate triage", "review dep PRs", "monthly deps", or on the 1st of a month if deps are grouped monthly.
use-class-for-state
For complex state: encapsulate multiple interacting variables, stateful algorithms, backtracking search state.