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.
git clone --depth 1 https://github.com/Yaomeng1749/claude-code-digital-nomad-config-packWrote 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/agents/yaomeng1749/claude-code-digital-nomad-config-pack/pytorch-build-resolver)<a href="https://agentmods.dev/agents/yaomeng1749/claude-code-digital-nomad-config-pack/pytorch-build-resolver"><img src="https://agentmods.dev/badge/agents/yaomeng1749/claude-code-digital-nomad-config-pack/pytorch-build-resolver/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/agents/yaomeng1749/claude-code-digital-nomad-config-pack/pytorch-build-resolver"><img src="https://agentmods.dev/badge/agents/yaomeng1749/claude-code-digital-nomad-config-pack/pytorch-build-resolver.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 | $0.00052 | $0.01335 |
| Opus 5 | $0.00026 | $0.00668 |
| Sonnet 5 | $0.00010 | $0.00267 |
| Haiku 4.5 | $0.00005 | $0.00134 |
Grade A, and why
pytorch-build-resolver 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 5d 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.
The source is not reproduced here
No licence file
A repository with no LICENSE is all rights reserved by default, so the body is not copied here. The metadata, the measurements and the link are.
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.
- 5d ago First seen · 121 lines · 52 tokens per session scan A 7ac671d451e7
pytorch-build-resolver is an agent published in the GitHub repository Yaomeng1749/claude-code-digital-nomad-config-pack (4 stars, last pushed 4mo ago), with no licence file. It adds 52 tokens to every session and 1,335 once invoked, about $0.0003 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 2026-09-03.
Other agents, from other repositories
build-error-resolver
A focused agent for restoring a failed software build with the smallest practical code changes.
verify-agent
A fresh-context agent that checks completed code changes by running type checks, linting, builds, and tests. Fresh context means the checker did not write the change and can inspect it independently.
pytorch-build-resolver
PyTorch runtime, CUDA, and training error resolution specialist. Fixes tensor shape mismatches, device errors, gradient issues, DataLoader problems, and mixed precision failures with minimal changes. Use when PyTorch training or inference crashes.
null-as-error-auditor
Read-only auditor for silent error swallowing in Effect code. Receives scanner JSON, reads source context, classifies severity, prescribes fixes. Spawned by /null-as-error skill.
error-boundary-architect
Use this agent when designing error handling strategy, building React error boundaries, creating fallback UIs, integrating error reporting (Sentry), or implementing graceful degradation. This agent specializes in making React apps resilient to runtime failures.
dc-lead
Use this agent for deep root-cause or multi-system investigation in a dev-crew run — stubborn, ambiguous, or circling failures that span subsystems, or a bugfix/debug loop that has already exceeded two passes. Reads the symptom and the codebase; produces a diagnosis and a contract the dev role can implement.