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 skills/gvkhosla/pi-tinker/tinker-debugnpx skills add gvkhosla/pi-tinker --skill tinker-debuggit clone --depth 1 https://github.com/gvkhosla/pi-tinkerWrote 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/gvkhosla/pi-tinker/tinker-debug)<a href="https://agentmods.dev/skills/gvkhosla/pi-tinker/tinker-debug"><img src="https://agentmods.dev/badge/skills/gvkhosla/pi-tinker/tinker-debug.svg" alt="Measured on agentmods" 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 | $0.00052 | $0.00475 |
| Opus 5 | $0.00026 | $0.00237 |
| Sonnet 5 | $0.00010 | $0.00095 |
| Haiku 4.5 | $0.00005 | $0.00047 |
Grade A, and why
tinker-debug 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 4d 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
Tinker debug (Pi overlay)
Do not duplicate Tinker Cookbook's debug skill. Start here, then load upstream.
First in Pi
/tinker doctor
That checks TINKER_API_KEY, Python 3.11+, Tinker SDK 0.23+, torch>=2.10, tml_renderers, Cookbook imports, retired model ids, and Inkling effort notes.
Install/upgrade:
uv pip install -U tinker-cookbook
There is no [inkling] extra. tml-renderers is a default Cookbook dependency.
Canonical triage
/plugin marketplace add thinking-machines-lab/tinker-cookbook
Then /tinker:debug. If the plugin is missing, read skills/debug/SKILL.md in thinking-machines-lab/tinker-cookbook.
Pi-specific tells
| Symptom | Check |
|---|---|
No matching distribution for tinker-cookbook[inkling] |
Extra removed. Install tinker-cookbook. |
ModuleNotFoundError: tml_renderers |
Reinstall Cookbook; needs Python 3.11+. |
TmlV0Renderer requires PyTorch 2.10 |
pip install "torch>=2.10". |
get_lr(...Inkling) NotImplementedError |
Expected. Calibrate LR; Cookbook publishes no default. |
| Train vs eval mismatch on Inkling | Effort not pinned. Same float for data, baseline, checkpoint. |
| Deployed Inkling to vLLM | Wrong. Stay on Tinker. See SERVING.md. |
| Export of DeepSeek/Kimi/gpt-oss as PEFT then vLLM LoRA | Cookbook merge path, not adapter serving. |
Inspect generated Python, metrics.jsonl, and checkpoints.jsonl. Those are the source of truth, not Pi UI state.
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.
- 4d ago First seen · 47 lines · 52 tokens per session scan A c1b7a3f3d297
tinker-debug is a skill published in the GitHub repository gvkhosla/pi-tinker (24 stars, last pushed 7d ago), licensed Apache-2.0. It adds 52 tokens to every session and 475 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 2026-08-30.
Other skills, from other repositories
fine-tuning-with-trl
Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.
fine-tuning-with-trl
Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.
Training
Guides fine-tuning and post-training work with explicit data, objective, hardware, and rollback assumptions.
fine-tuning-with-trl
Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.
model_finetuning
Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.
fine-tuning-with-trl
Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.