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 infiniV/ultra-ml-intern --skill groundinggit clone --depth 1 https://github.com/infiniV/ultra-ml-internWrote 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/infiniv/ultra-ml-intern/grounding)<a href="https://agentmods.dev/skills/infiniv/ultra-ml-intern/grounding"><img src="https://agentmods.dev/badge/skills/infiniv/ultra-ml-intern/grounding/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/infiniv/ultra-ml-intern/grounding"><img src="https://agentmods.dev/badge/skills/infiniv/ultra-ml-intern/grounding.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 | $0.00147 | $0.01533 |
| Opus 5 | $0.00073 | $0.00766 |
| Sonnet 5 | $0.00029 | $0.00307 |
| Haiku 4.5 | $0.00015 | $0.00153 |
Grade A, and why
grounding 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 9d 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 — 145 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Grounding
model-provenance harvests a model's real code, papers, and checkpoint
metadata once. This skill is what every later session does with it: read the
archived source before writing code, so the checkpoint ids, class names,
preprocessing constants, and recipe come from the model's actual source rather
than from recall.
That distinction is the whole point. Your memory of a specific model is plausible and lossy in a way that is hard to notice: checkpoint ids that look exactly right and 404, a class name borrowed from a neighbouring model, a normalization constant from the previous version, an argument that was renamed a release ago. All of it reads as confident, correct code. The archive is the only thing on this machine that can settle those questions, so consult it first and check against it after.
Workflow
1. Find the archive
scripts/load_archive.py "<model name>"
Name matching is fuzzy — DINOv3, dinov3, and "SAM 2" all resolve. It
prints the archive path, notes.md's outline, the captured checkpoint ids with
their gated/license status, and what the archive is missing.
scripts/load_archive.py --list shows every archive on this machine.
If there is no archive, this skill has nothing to offer. Say so plainly and pick one:
- Run
model-provenanceto harvest the model first — right when the code matters, or the model is obscure, or a wrong checkpoint id costs a long download. - Proceed from memory, having told the user that the model's details are ungrounded and that checkpoint ids and API signatures may be wrong.
Never quietly fall back to memory. An ungrounded answer that looks identical to a grounded one is the failure this skill exists to prevent.
2. Read before writing
Read notes.md first, in full. It is the synthesis: architecture, training
recipe, variant table, I/O contract, versions and license, a working inference
snippet, and known gotchas — every claim cited to a file and line.
Then read only what the task needs, following notes.md's citations:
What ships with it
2 files 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.
- 9d ago First seen · 145 lines · 147 tokens per session scan A d8277dc59bba
grounding is a skill published in the GitHub repository infiniV/ultra-ml-intern (3 stars, last pushed 22d ago), licensed MIT. It adds 147 tokens to every session and 1,533 once invoked, about $0.0007 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 2026-08-31.
Other skills, from other repositories
prompt-engineering
Create, update, review, or discuss an LLM prompt — a system prompt, a skill, or an agent. Use when writing or improving a prompt, discussing a skill or agent, diagnosing prompt failures, or when the user says a prompt needs work.
review-prompt
Review LLM prompts against the prompt-engineering skill's principles — leading with where each line came from — and report the findings without modifying files. Use when reviewing prompt quality, auditing a prompt, evaluating a system prompt, or checking whether prompt issues are high-confidence and fixable.
ccc-prompt-fix
Fix and sharpen a prompt. Diagnoses it against the 6 prompt-quality patterns, returns a tightened rewrite with the reasoning, and suggests the right library prompt for your task.
eval-set
Build and run a project-specific retrieval eval so changes to your rules or base prompt are scored, not eyeballed. Mirrors tests/groundtruth.json + clawness eval: you write prompt→expected-rule cases, then measure MRR@k and hit-rate before and after an edit. Run it after trimming a base prompt into ranked retrieval…
oracle-grill-me
Grill the user relentlessly about a plan, decision, or idea, maintaining the project's domain model (CONTEXT.md, ADRs) as decisions crystallise. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.
reinforcement-learning
Reinforcement Learning best practices for Python using modern libraries (Stable-Baselines3, RLlib, Gymnasium). Use when: Implementing RL algorithms (PPO, SAC, DQN, TD3, A2C) Creating custom Gymnasium environments Training, debugging, or evaluating RL agents Setting up hyperparameter tuning for RL Deploying RL models…