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 MadGraphTeam/MadAgents --skill mg-probegit clone --depth 1 https://github.com/MadGraphTeam/MadAgentsWrote 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/madgraphteam/madagents/mg-probe)<a href="https://agentmods.dev/skills/madgraphteam/madagents/mg-probe"><img src="https://agentmods.dev/badge/skills/madgraphteam/madagents/mg-probe/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/madgraphteam/madagents/mg-probe"><img src="https://agentmods.dev/badge/skills/madgraphteam/madagents/mg-probe.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 | $0.00118 | $0.01083 |
| Opus 5.5 | $0.00047 | $0.00433 |
| Sonnet 5.5 | $0.00024 | $0.00217 |
| Haiku 4.5 | $0.00012 | $0.00108 |
Grade A, and why
mg-probe 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.
How it starts
The opening of the file, as written. The whole thing — 45 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/mg-probe
Source-walking by consultants is the load-bearing verification. The probe is the runtime check, dispatched adversarially: given a target state of what should be true if the spec is correct, the probe finds where running MadGraph deviates from it.
Frame the dispatch as "find what's wrong with running this," not "confirm the consultant claims." Adversarial framing produces different probes — they try to break expectations, including ones the dispatch did not name.
Writing well-formed expectations
Expectations come in two kinds, and the list must carry at least one of the second:
- Spec-derived — what the assembled spec implies should happen ("this
generateline implies N diagrams"). A spec-derived match only confirms the spec is internally what it claims; a spec that parses and runs but computes the wrong observable still matches its own spec-derived expectations. - Target-derived — what a correct setup for the question produces, computed from the physics of the request independently of the assembled spec: the expected signal topology / diagram count for the requested process, the order-of-magnitude σ for the requested observable, non-zero-ness in the kinematically allowed region. The physics regime classification is the natural source. A target-derived expectation is what makes a wrong-but-runnable spec deviate — a contaminated form whose σ is orders of magnitude off the physics target contradicts the anchor even though it matches its own spec.
Label each expectation spec-derived or target-derived; a list with only spec-derived expectations tests nothing a wrong spec would fail.
Plain-prose claims about what MadGraph should do, observable by running, file inspection, grep, or launch:
- "
generate p p > h j j QCD=0, (h > w+ w-, w+ > ta+ vt, w- > ta- vt~)should produce 87 diagrams, NEXTERNAL=8, IDUP showing tau+ nu_tau tau- nu_tau~, and gForceBW=1 for both W± propagators in decayBW.inc." - "No 'decay discarded' or 'no phase space' warnings in stdout."
- "matrix1_orig.f contains zero Yukawa-Hττ vertex calls (no
FFS4.*GC_99.*MDL_MHhits) — Yukawa channels excluded by chain syntax." - "At bwcutoff=50 with MH=100, σ is non-zero — the BW window covers the W* phase space."
- "The integrator survives
launchwithblock smeft 35 1.0and-masslessrestriction defaults — no multi-channel crash."
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 · 45 lines · 118 tokens per session scan A 42dd635a9bd8
mg-probe is a skill published in the GitHub repository MadGraphTeam/MadAgents (14 stars, last pushed 2mo ago), licensed MIT. It adds 118 tokens to every session and 1,083 once invoked, about $0.0005 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 2026-09-03.
Other skills, from other repositories
train-pose
Train/evaluate WiFi pose models honestly — camera-supervised (MediaPipe + CSI) and camera-free (WiFlow), always checked against the mean-pose baseline before any PCK is quoted.
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
experiment-audit
A review step that checks whether an experiment's reported results are supported by real data and a sufficiently broad test.
i4h-catheter-navigation-e2e
End-to-end smoke for catheter navigation covering setup, digital twin, DRR, and unit tests. Use when asked to run the full catheter workflow smoke or demo the v0.7 pipeline.
i4h-catheter-navigation-render-drr
Render a single DRR fluoroscopy frame from a CT cache or synthetic phantom. Use when asked to render DRR, generate a fluoro image, or smoke-test the Slang renderer.
i4h-catheter-navigation-smoke
Run CPU-only fluorosim smoke tests (imports, preprocessing, CLI parsers). Use when asked to smoke-test catheter navigation in CI or without a GPU.