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 rules/amdresearch/ai4science-studio/cross-model-lesson-reusegit clone --depth 1 https://github.com/AMDResearch/ai4science-studioWrote 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/rules/amdresearch/ai4science-studio/cross-model-lesson-reuse)<a href="https://agentmods.dev/rules/amdresearch/ai4science-studio/cross-model-lesson-reuse"><img src="https://agentmods.dev/badge/rules/amdresearch/ai4science-studio/cross-model-lesson-reuse.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 5.1 | $0.00619 | $0.00619 |
| Opus 5 | $0.00309 | $0.00309 |
| Sonnet 5 | $0.00124 | $0.00124 |
| Haiku 4.5 | $0.00062 | $0.00062 |
Grade A, and why
cross-model-lesson-reuse 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 6d 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 — 34 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Organic lesson capture and propagation
Every bug fix, workaround, or pattern discovery is a lesson. Do not treat skill/rule/recipe updates as a separate task. Fold them into the fix itself.
When you fix anything, do ALL of these in the same pass
- Fix the immediate script. Apply the fix to the file that failed.
- Propagate to sibling models. Scan
*/models/*/examples/for the same pattern. Fix them now, not later. - Propagate across runtimes. If the fix was in an Apptainer script, check the Docker equivalent (and vice versa). Common cross-runtime issues:
- SCRIPT_DIR resolution, read-only FS handling, dep lists, env var clobbering, torch protection
- Update the relevant SKILL.md. Add the lesson to the right skill file:
- Repo-wide →
.cursor/skills/ai4science-studio/SKILL.md - Domain-specific →
.cursor/skills/ai4science-<domain>/SKILL.md
- Repo-wide →
- Update or create a rule if the lesson is a recurring process mistake (not just a technical fix).
The test: would a future session make the same mistake?
If you fixed something and a fresh agent session working on a different model could hit the same issue, you haven't finished. The fix must be discoverable in a skill or rule before the task is done.
Examples
| Discovery | Immediate fix | Must also do |
|---|---|---|
--no-deps drops transitive runtime deps |
Switch to --extra-index-url + strip for StormCast |
Apply same approach to ORBIT-2 install-to-tmp; update studio SKILL.md to say "prefer --extra-index-url + strip over --no-deps for runtime installs" |
| SCRIPT_DIR resolves to SLURM spool | Add scontrol trick to StormCast |
Add to ORBIT-2, GP-MoLFormer, any future model; already in studio SKILL.md — verify it's there |
run_generation.sh has CRLF endings |
sed -i 's/\r$//' |
Check all .sh files in the repo; add a note to SKILL.md about verifying line endings |
| SIF is read-only, pip install fails | Add install-to-tmp + bind-mount | Apply to every model that does inline pip install in Apptainer; update SKILL.md with the pattern |
rdkit-pypi has no cp312 wheels |
Graceful fallback with || echo |
Note py3.12 incompatibility in healthcare SKILL.md; check other models for similar version-gated deps |
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.
- 6d ago First seen · 34 lines · 619 tokens per session scan A 350df71d8b2e
cross-model-lesson-reuse is a cursor rule published in the GitHub repository AMDResearch/ai4science-studio (4 stars, last pushed 1mo ago), licensed MIT. It adds 619 tokens to every session, about $0.0031 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 cursor rules, from other repositories
ponytail
Ponytail, lazy senior dev mode. Always pick the simplest solution that works.
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.
dev-standard
Apache Superset development standards and guidelines for Cursor IDE.
cli-error-handling
CLI command error handling patterns.
control-plane-descriptors
Control plane descriptor and instance implementation patterns.
family-instance-domain-actions
Family instance domain action implementation patterns.