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 DeepLink-org/probing --skill training_hanggit clone --depth 1 https://github.com/DeepLink-org/probingWrote 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/deeplink-org/probing/training_hang)<a href="https://agentmods.dev/skills/deeplink-org/probing/training_hang"><img src="https://agentmods.dev/badge/skills/deeplink-org/probing/training_hang/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/deeplink-org/probing/training_hang"><img src="https://agentmods.dev/badge/skills/deeplink-org/probing/training_hang.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 | $0.00015 | $0.00267 |
| Opus 5.5 | $0.00006 | $0.00107 |
| Sonnet 5.5 | $0.00003 | $0.00053 |
| Haiku 4.5 | $0.00002 | $0.00027 |
Grade A, and why
training_hang 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.
What it actually says
Training hang diagnosis
Loss 不更新、进程仍存活但无进展时使用。 组合：实时混合栈 + CPU 线程 wchan + 最近 trace 活动。
Parameters
stack_depth(integer, default20): Backtrace frames to show
Related skills
- 栈在 dist/barrier → 检查各 rank 是否都存活: probing list / cluster nodes
- 栈在 DataLoader → 减少 num_workers 或检查数据路径
- 需要持续采样 → SET probing.pprof.sample_freq=99 后 skill: module_bottleneck
What ships with it
1 file 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.
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 · 29 lines · 15 tokens per session scan A 8668365baf1e
training_hang is a skill published in the GitHub repository DeepLink-org/probing (10 stars, last pushed 2d ago), licensed Apache-2.0. It adds 15 tokens to every session and 267 once invoked, about $0.0001 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 2026-08-30.
Other skills, from other repositories
notebooklm
Install, authenticate, troubleshoot, and operate Gemini Notebook through the notebooklm-py CLI or typed async Python API. Use for notebook and source management, grounded chat and research, and artifact generation or download when the user mentions Gemini Notebook, notebooklm-py, the notebooklm CLI, or its Python API.…
exploring-llm-traces
Debug and inspect LLM/AI agent traces using PostHog's MCP tools. Use when the user pastes a trace or session URL (e.g. /ai-observability/traces/ or /ai-observability/sessions/ ), asks to debug a trace, figure out what went wrong, check if an agent used a tool correctly, verify context/files were surfaced, inspect…
code-runner
A sandboxed code execution tool that runs short programs without using your computer's files or network. It supports Python, JavaScript, TypeScript, Go, Rust, Java, C, C++, Ruby, and PHP.
art-of-debugging
Systematic methodology and concrete tool recipes for debugging Unix, Python, and PyTorch programs - crashes, hangs, segfaults, wrong output, CUDA OOM, NaN/Inf, slowness, and multi-node/multi-GPU issues. Use when a program crashes, hangs, deadlocks, segfaults, runs out of memory (OOM), produces NaN/Inf or wrong…
operate-ai-stack
Develop, test, and troubleshoot this generated AI or ML workload across its framework, model and embedding providers, retrieval stores, interfaces, training, serving, and observability layers. Use when changing prompts, tools, agents, RAG, MCP, inference, fine-tuning, evaluations, or provider integration.
inference-precision-tensor-dump-compare
A debugging tool that captures intermediate tensors, the numerical data passed between model layers, during inference and compares them between GPUs and Ascend NPUs. It supports vLLM and SGLang inference frameworks.