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 TianGzlab/OmicsClaw --skill spatial-deconvgit clone --depth 1 https://github.com/TianGzlab/OmicsClawWrote 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/tiangzlab/omicsclaw/spatial-deconv)<a href="https://agentmods.dev/skills/tiangzlab/omicsclaw/spatial-deconv"><img src="https://agentmods.dev/badge/skills/tiangzlab/omicsclaw/spatial-deconv/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/tiangzlab/omicsclaw/spatial-deconv"><img src="https://agentmods.dev/badge/skills/tiangzlab/omicsclaw/spatial-deconv.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
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 →
- high Rogue Agent · line 3 Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.Fix: Prevent the skill from modifying its own code, SKILL.md, or configuration files. Treat skill files as read-only at runtime.
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 | $0.00084 | $0.01384 |
| Opus 5.5 | $0.00034 | $0.00554 |
| Sonnet 5.5 | $0.00017 | $0.00277 |
| Haiku 4.5 | $0.00008 | $0.00138 |
Grade A, and why
spatial-deconv 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 17d 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 — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.
spatial-deconv
When to use
Estimate cell-type proportions for spatial spots using a labelled single-cell reference. Use spatial-annotate for discrete labels and spatial-domains for tissue regions. Every method requires a reference; the CLI default is cell2location.
Use from a step
from skills._sdk.notebook import load_skill, read_input, write_output
spatial = read_input("spatial.h5ad")
reference = read_input("reference.h5ad")
library = load_skill("spatial-deconv")
library.deconvolve(spatial, reference=reference, method="flashdeconv", random_state=0)
write_output(library.proportions(spatial), "tables/proportions.csv")
examples/example_step.py uses an explicit synthetic reference and verifies known stripe identities; it is not a real-tissue benchmark. It sets lambda_spatial=10 for the small grid; the CLI default remains 5000.
API
deconvolve(adata, *, reference, method: str='cell2location', cell_type_key: str='cell_type', random_state: int=0, **parameters)
Estimate proportions in place and return the same spatial AnnData.
Count models read layers['counts'], raw or X in that order. FlashDeconv, Tangram and SPOTlight read nonnegative X. Reference data are copied. RCTD, SPOTlight and CARD retain their temporary R bridges; their wrappers do not expose a seed and results vary between runs.
:param adata: Spatial AnnData; input expression is retained. :param reference: Required labelled reference AnnData; load it with read_input. :param method: CLI default cell2location; flashdeconv, rctd, destvi, stereoscope, tangram, spotlight and card are also supported. :param cell_type_key: Reference label column, CLI default cell_type. :param random_state: Seed for FlashDeconv/Tangram and scvi training, default 0. :param parameters: Method options in references/parameters.md, with CLI prefixes removed; omitted options keep their CLI defaults. :returns: The same AnnData with deconvolution_, labels and JSON diagnostics. :raises ValueError: Invalid method, expression, reference labels or backend proportions. :raises TypeError: Reference is not AnnData or a method option is unknown. :raises ImportError: A backend is missing; use install_skill_deps for its named package.
What ships with it
10 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.
- _api.py 7.1 KB runs code
- examples/example_step.py 880 B runs code
- r_visualization/deconv_publication_template.R 5.3 KB
- r_visualization/README.md 1.6 KB
- references/methodology.md 7.1 KB
- references/output_contract.md 2.0 KB
- references/parameters.md 8.1 KB
- spatial_deconv.py 56 KB runs code
- tests/test_api.py 1.3 KB runs code
- tests/test_spatial_deconv.py 10 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.
- 17d ago Changed · -30 lines db63f1bff7d1
- 1mo ago First seen · 168 lines · 84 tokens per session scan A dd4ae645dc13
spatial-deconv is a skill published in the GitHub repository TianGzlab/OmicsClaw (161 stars, last pushed 17d ago), licensed Apache-2.0. It adds 84 tokens to every session and 1,384 once invoked, about $0.0003 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
spatial-deconvolution
Cell-type deconvolution for spatial transcriptomics by mapping a single-cell RNA-seq reference onto spatial data.
spatial-omics
Spatial transcriptomics and spatial proteomics analysis covering technology-specific workflows, spatial statistics, deconvolution, and niche analysis.
spatial-cell-annotation
Cell type annotation for spatial transcriptomics data using marker-based scoring, Tangram mapping, scANVI transfer, or CellAssign probabilistic models.
spatial-sc2spatial
Python API skill for mapping reference single-cell annotations and expression programs onto spatial transcriptomics data.
onekgpd
Queries the 1000 Genomes Project dataset (3,202 whole-genome-sequenced individuals, GRCh38) at the level of individual participants. Use when a question is about individuals or variants in the 1000 Genomes Project cohort: which individuals carry variants matching specific criteria in a gene or region, which…
neuropixels-analysis
Analyzes Neuropixels extracellular recordings end-to-end with SpikeInterface. Covers loading SpikeGLX/Open Ephys/NWB data, preprocessing, drift/motion correction, Kilosort4 (and CPU) spike sorting, quality metrics, and unit curation (threshold-based, model-based UnitRefine, and AI-assisted visual review). Use when…