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 zjunlp/Mechanist --skill mechanism-skillsgit clone --depth 1 https://github.com/zjunlp/MechanistWrote 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/zjunlp/mechanist/mechanism-skills)<a href="https://agentmods.dev/skills/zjunlp/mechanist/mechanism-skills"><img src="https://agentmods.dev/badge/skills/zjunlp/mechanist/mechanism-skills/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/zjunlp/mechanist/mechanism-skills"><img src="https://agentmods.dev/badge/skills/zjunlp/mechanist/mechanism-skills.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
SkillSpector: 1 finding, up to low
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- low Excessive Agency · line 41 Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.Fix: Limit the skill's scope to its documented purpose. Remove instructions that enable the agent to perform actions outside its stated functionality.
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 | $0.00214 | $0.06965 |
| Opus 5 | $0.00107 | $0.03483 |
| Sonnet 5 | $0.00043 | $0.01393 |
| Haiku 4.5 | $0.00021 | $0.00696 |
Grade A, and why
mechanism-skills 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 11d 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 — 260 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Mechanism Skills
A curated set of eleven method families for analyzing the internal mechanisms of neural networks. Use this document as the routing entry point: read it first, pick the family that matches the question, then follow the link into the sub-skill for submethod detail, demos, and reference implementations.
When to Use
Consult these skills whenever the research question is about how the model computes internally, not just how it performs externally. Typical triggers:
- A reviewer asks for mechanistic evidence behind a claim
- A claim attributes a behavior to a specific component (layer, head, neuron, feature, weight)
- An experiment needs to isolate which internal object drives a behavior
- A method swap is under consideration and the candidate pool should include mechanism-level tools (e.g., swapping a magnitude-based screen for a causal-intervention method)
- A vision or vision-language model needs concept-level explanations rather than raw saliency
If the research question is purely behavioral (accuracy, robustness, calibration) and does not depend on internal structure, these skills are not the right tool — continue with the standard experiment skills.
Loading Protocol (Mandatory)
Skills cascade strictly top-down: this routing file → family SKILL.md → submethod SKILL.md. Each level adds detail the level above only sketches. Never act on a summary from a higher level alone.
Hard requirements — these are not suggestions:
- Pick a family based on the research question (use the Selecting a Method table). The summaries below are previews only — they are intentionally insufficient to plan or run work.
- Once a family is selected, you MUST load
<family>/SKILL.mdin full before any further reasoning, recommendation, or code in that family. Assumed objects, exact signal definitions, scope conditions, composition rules across submethods, and known failure modes live in the family file, not here. - Once a submethod is selected, you MUST load
<family>/<submethod>/SKILL.mdin full before running, citing, choosing hyperparameters for, or recommending that submethod. Runnable scripts, API conventions, defaults, and submethod-specific gotchas live there. - Do not skip levels. Going from this file directly to a submethod without reading the family SKILL.md is forbidden, because cross-submethod composition (screen-then-verify pairings, shared metrics, layer/position conventions) is defined at the family level.
- Re-load on switch. If the analysis pivots to a different family or submethod, repeat the cascade for the new branch — do not carry assumptions across families.
What ships with it
60 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.
- causal-attribution/ablation/references/api_reference.md 8.6 KB
- causal-attribution/ablation/scripts/intervention_experiments.py 14 KB runs code
- causal-attribution/ablation/scripts/model_analysis.py 12 KB runs code
- causal-attribution/ablation/SKILL.md 26 KB
- causal-attribution/article_references.md 430 B
- causal-attribution/attribution-patching/references/api_reference.md 7.1 KB
- causal-attribution/attribution-patching/scripts/ioi_dataset_usage.py 12 KB runs code
- causal-attribution/attribution-patching/scripts/run_attribution_patching.py 11 KB runs code
- causal-attribution/attribution-patching/SKILL.md 24 KB
- causal-attribution/patching/references/rome_api_reference.md 9.0 KB
- causal-attribution/patching/scripts/causal_tracing_demo.py 11 KB runs code
- causal-attribution/patching/scripts/rome_editing_example.py 8.4 KB runs code
- causal-attribution/patching/SKILL.md 20 KB
- causal-attribution/SKILL.md 3.5 KB
- circuit-discovery/article_references.md 334 B
- circuit-discovery/attribution-based-edge-scoring/article_references.md 165 B
- circuit-discovery/attribution-based-edge-scoring/references/api_reference.md 4.8 KB
- circuit-discovery/attribution-based-edge-scoring/scripts/usage_example.py 2.7 KB runs code
- circuit-discovery/attribution-based-edge-scoring/SKILL.md 3.1 KB
- circuit-discovery/intervention-based-edge-search/article_references.md 147 B
- circuit-discovery/intervention-based-edge-search/references/api_reference.md 5.2 KB
- circuit-discovery/intervention-based-edge-search/scripts/acdc_run_demo.py 1.4 KB runs code
- circuit-discovery/intervention-based-edge-search/SKILL.md 1.7 KB
- circuit-discovery/SKILL.md 4.3 KB
- feature-dictionary-learning/article_references.md 636 B
- feature-dictionary-learning/crosscoder/article_references.md 164 B
- feature-dictionary-learning/crosscoder/scripts/train_and_evaluate_sae.py 10 KB runs code
- feature-dictionary-learning/crosscoder/scripts/train_sae_demo.py 9.7 KB runs code
- feature-dictionary-learning/crosscoder/SKILL.md 23 KB
- feature-dictionary-learning/ica-lens/article_references.md 154 B
- feature-dictionary-learning/ica-lens/references/api_reference.md 8.9 KB
- feature-dictionary-learning/ica-lens/scripts/artifact_variant_example.py 3.5 KB runs code
- feature-dictionary-learning/ica-lens/scripts/explorer_app_usage.py 4.2 KB runs code
- feature-dictionary-learning/ica-lens/SKILL.md 20 KB
- feature-dictionary-learning/SAE/article_references.md 150 B
- feature-dictionary-learning/SAE/SKILL.md 3.0 KB
- feature-dictionary-learning/SKILL.md 6.5 KB
- feature-dictionary-learning/transcoder/article_references.md 130 B
- feature-dictionary-learning/transcoder/references/api_reference.md 12 KB
- feature-dictionary-learning/transcoder/scripts/train_transcoder_example.py 5.8 KB runs code
- feature-dictionary-learning/transcoder/scripts/transcoder_usage_example.py 9.0 KB runs code
- feature-dictionary-learning/transcoder/SKILL.md 22 KB
- gradient-detection/article_references.md 482 B
- gradient-detection/inputs-and-layer-wise-states/references/api_reference.md 2.3 KB
- gradient-detection/inputs-and-layer-wise-states/scripts/calculate_gradients.py 8.4 KB runs code
- gradient-detection/inputs-and-layer-wise-states/scripts/visualize_gradients.py 13 KB runs code
- gradient-detection/inputs-and-layer-wise-states/SKILL.md 22 KB
- gradient-detection/intermediate-outputs/references/relp_api_reference.md 9.1 KB
- gradient-detection/intermediate-outputs/scripts/basic_relp_analysis.py 9.3 KB runs code
- gradient-detection/intermediate-outputs/scripts/ioi_task_analysis.py 14 KB runs code
- gradient-detection/intermediate-outputs/SKILL.md 23 KB
- gradient-detection/parameters/references/api_reference.md 8.1 KB
- gradient-detection/parameters/scripts/extract_linguistic_regions.py 16 KB runs code
- gradient-detection/parameters/scripts/preprocess_multilingual_data.py 9.1 KB runs code
- gradient-detection/parameters/SKILL.md 26 KB
- gradient-detection/SKILL.md 3.9 KB
- magnitude-analysis/article_references.md 594 B
- magnitude-analysis/dynamic-components/references/api_reference.md 6.8 KB
- magnitude-analysis/dynamic-components/scripts/compute_perplexity.py 15 KB runs code
- magnitude-analysis/dynamic-components/scripts/deactivate_neurons.py 13 KB runs code
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.
- 11d ago First seen · 260 lines · 214 tokens per session scan A 7cc9cbaec01d
mechanism-skills is a skill published in the GitHub repository zjunlp/Mechanist (74 stars, last pushed 15d ago), licensed MIT. It adds 214 tokens to every session and 6,965 once invoked, about $0.0011 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
nanoresearch-experiment
Generate a Python code skeleton from an experiment blueprint.
nanoresearch-writing
Draft a LaTeX research paper from all previous stage outputs.
nanoresearch-ideation
Search academic literature and generate research hypotheses.
nnsight-remote-interpretability
Provides guidance for interpreting and manipulating neural network internals using nnsight with optional NDIF remote execution. Use when needing to run interpretability experiments on massive models (70B+) without local GPU resources, or when working with any PyTorch architecture.
sparse-autoencoder-training
Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network activations into interpretable features. Use when discovering interpretable features, analyzing superposition, or studying monosemantic representations in language models.
transformer-lens-interpretability
Provides guidance for mechanistic interpretability research using TransformerLens to inspect and manipulate transformer internals via HookPoints and activation caching. Use when reverse-engineering model algorithms, studying attention patterns, or performing activation patching experiments.