← Back to Skills Marketplace
billwanttobetop

Auto Proteomics

by Billwanttobetop · GitHub ↗ · v0.2.0 · MIT-0
cross-platform ✓ Security Clean
123
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install auto-proteomics
Description
Public OpenClaw skill for low-token routing and downstream analysis of processed DDA LFQ proteomics inputs. Use when the user already has protein-level quant...
README (SKILL.md)

Auto Proteomics

Author: Guo Xuan 郭轩
Contact: [email protected]

auto-proteomics is a public v0.x skill for processed proteomics downstream work.

The current public promise is intentionally narrow:

  • one shipped runnable workflow: dda-lfq-processed
  • one public input family: processed DDA LFQ protein-level tables
  • one public comparison model: group-a vs group-b

Everything else in this repository should be read as routing context, internal prototype, or future scaffold unless a document explicitly marks it as part of the public promise. Presence of a script, schema, or branch document does not mean the route is publicly supported. In particular, dia-quant is intentionally exposed as an internal prototype route for correct routing and contract validation, not as a shipped public workflow.

Use this skill when

  • the user already has processed protein-level quantification output
  • the main table is MaxQuant-like proteinGroups.txt
  • the goal is QC, normalized matrix generation, and two-group differential protein analysis
  • the user wants a low-token, file-driven workflow instead of a long chat-only protocol

Do not use this skill when

  • the user starts from raw spectra and needs search/identification
  • the request is primarily DIA, phosphoproteomics, enrichment, or multi-omics execution
  • the task requires more than one comparison design in the current release
  • the user only wants generic statistics with no proteomics context

Public promise in v0.x

Shipped and supported now:

  • route processed DDA LFQ downstream requests into dda-lfq-processed
  • validate the expected processed-input shape
  • generate matrix, QC, differential tables, report, and manifest outputs

Not promised yet:

  • raw-spectrum search pipelines
  • DIA public execution support
  • phosphoproteomics execution
  • enrichment execution
  • multi-omics execution
  • generalized study-design handling beyond the current two-group path

Internal prototype route available for routing only:

  • dia-quant may be selected only when the request is explicitly about processed DIA quant tables that fit the checked-in DIA contract
  • selecting dia-quant means internal prototype triage, never a public v0.x execution recommendation

Important boundary:

  • non-shipped branches may contain scaffold or prototype execution files for internal framework development
  • smaller models must not treat those files as public runnable recommendations unless a route is explicitly marked shipped

Minimal workflow

  1. Read references/WORKFLOW_INDEX.yaml
  2. If the route is unclear, run scripts/decision/route_proteomics.py
  3. Check that the request fits the public v0.x boundary
  4. Run scripts/workflows/dda_lfq_processed.sh
  5. Use references/ for runtime, onboarding, and development rules

Public runnable entrypoint

bash scripts/workflows/dda_lfq_processed.sh \
  --input-dir \x3Crun_dir> \
  --protein-groups \x3CproteinGroups.txt> \
  --summary \x3Csummary.txt> \
  --parameters \x3Cparameters.txt> \
  --output-dir \x3Coutput_dir> \
  --group-a \x3Ccondition_a> \
  --group-b \x3Ccondition_b>

Input contract

Required:

  • proteinGroups.txt with LFQ intensity * or Intensity * columns
  • summary.txt with Raw file and Experiment columns

Optional:

  • parameters.txt

Output contract

The shipped workflow produces:

  • normalized protein matrix files under matrix/
  • QC outputs under qc/
  • differential protein tables under stats/
  • REPORT.md
  • summary.json
  • run_manifest.json

Repository layers

  • SKILL.md: public entry and release boundary
  • references/WORKFLOW_INDEX.yaml: machine-readable routing and shipped-vs-non-shipped map
  • references/BRANCH_FRAMEWORK.md: standard branch contract for future routes
  • references/branches/: per-branch specs for scaffold and prototype workflows
  • references/DIA_INPUT_SCHEMA.md: first narrow schema for DIA prototype intake
  • scripts/workflows/dda_lfq_processed.sh: shipped workflow entrypoint
  • shipped public guidance lives in documents that explicitly describe the processed DDA v0.x path
  • non-shipped reference docs exist for internal framework development and must not be surfaced as public support

Read next

  • references/WORKFLOW_INDEX.yaml
  • references/RUNTIME_REQUIREMENTS.md
  • references/BRANCH_FRAMEWORK.md
  • references/DEMO_INPUT_GUIDE.md
  • references/DEVELOPMENT_GUIDE.md
Usage Guidance
This repository appears coherent and limited to processed DDA LFQ downstream analysis. Before installing or running it: (1) inspect the included scripts (shell/Python) yourself or run them in an isolated environment to confirm they do only local file processing and do not perform unexpected network access; (2) ensure your environment has bash, python3, and PyYAML as documented (the package metadata omitted these); (3) provide only the intended processed input files (proteinGroups.txt, summary.txt) — do not feed raw vendor files expecting the skill to perform raw-spectrum searches; (4) if you will run it inside an agent that can invoke skills autonomously, be aware the agent could run the shipped workflow automatically when given matching inputs — this is normal but worth acknowledging. If you need deeper assurance, request the full contents of the main scripts to review for any network or credential usage before running on sensitive data.
Capability Analysis
Type: OpenClaw Skill Name: auto-proteomics Version: 0.2.0 The 'auto-proteomics' skill bundle is a legitimate bioinformatics toolkit designed for downstream analysis of proteomics data. The code consists of Bash workflow wrappers and Python scripts for data standardization, matrix construction, quality control, and differential expression analysis (e.g., `dda_lfq_processed.sh`, `differential_protein.py`, and `qc_filter_normalize.py`). The instructions in SKILL.md and the routing logic in `route_proteomics.py` are focused on guiding the AI agent to the correct analysis path based on user input, with explicit boundaries between supported and prototype features. No evidence of data exfiltration, malicious execution, or harmful prompt injection was found.
Capability Assessment
Purpose & Capability
Name, description, and included scripts match the stated goal: routing validated processed DDA LFQ inputs into a shipped two-group downstream workflow. No unrelated credentials, binaries, or external services are requested.
Instruction Scope
SKILL.md gives a focused, limited runtime procedure (validate routing, then run scripts/workflows/dda_lfq_processed.sh). Instructions do not ask the agent to read unrelated system files or to transmit data to external endpoints. The repository explicitly documents public vs prototype branches to avoid over-promising.
Install Mechanism
This is an instruction-only skill with no install spec (low risk). Minor inconsistency: the skill metadata lists 'no required binaries', but the runtime docs (references/RUNTIME_REQUIREMENTS.md and SKILL.md) expect bash, python3 and PyYAML. That is a documentation/packaging mismatch to be aware of but not an active supply-chain risk.
Credentials
No environment variables, credentials, or config paths are required by the skill. The requested surface is proportionate to the stated proteomics processing task.
Persistence & Privilege
The skill is not always-enabled and is user-invocable; model invocation is allowed (platform default). This autonomous-invocation ability is normal — there are no other high-privilege behaviors requested.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install auto-proteomics
  3. After installation, invoke the skill by name or use /auto-proteomics
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.2.0
v0.2.0: tighter honest public boundary, bilingual release materials, stricter package hygiene, and internal DIA prototype routing/validation groundwork.
v0.1.0
Initial public release: processed DDA LFQ downstream workflow with clear v0.x boundary, runnable execution files, routing index, runtime docs, and lightweight demo inputs.
Metadata
Slug auto-proteomics
Version 0.2.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 2
Frequently Asked Questions

What is Auto Proteomics?

Public OpenClaw skill for low-token routing and downstream analysis of processed DDA LFQ proteomics inputs. Use when the user already has protein-level quant... It is an AI Agent Skill for Claude Code / OpenClaw, with 123 downloads so far.

How do I install Auto Proteomics?

Run "/install auto-proteomics" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Auto Proteomics free?

Yes, Auto Proteomics is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Auto Proteomics support?

Auto Proteomics is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Auto Proteomics?

It is built and maintained by Billwanttobetop (@billwanttobetop); the current version is v0.2.0.

💬 Comments