Skip to content
Back to search

science-skills

This repository contains the prompt, skill, agent, and MCP server assets for a Claude Science style scientific computing assistant. The contents are mostly runtime-facing configuration and…

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/jimliu--science-skills/badge.svg)](/listings/jimliu--science-skills)

HTML

<a href="/listings/jimliu--science-skills"><img src="/api/listings/jimliu--science-skills/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 131 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. # Science Skills
  2. This repository contains the prompt, skill, agent, and MCP server assets for a
  3. Claude Science style scientific computing assistant. The contents are mostly
  4. runtime-facing configuration and documentation: skills teach the assistant how
  5. to perform domain workflows, agents define specialized profiles, and MCP
  6. servers expose scientific data and chemistry tools.
  7. ## Source
  8. The `skills/` and `mcp-servers/` assets match **Claude Science 0.1.71**
  9. (public release, macOS arm64, built October 26, 2026). They are copied directly
  10. from the app's embedded asset archive, including helper code, dependencies,
  11. licenses, and widget assets.
  12. `VERSION` records this asset snapshot version. `SOURCE.json` records the app's
  13. build metadata, source checksums, and file-tree checksums for both directories.
  14. Upstream package manifests and skill metadata are preserved verbatim; their
  15. dependency versions are independent of the app version. The base prompt and
  16. agent profiles are retained from the earlier repository snapshot and are not
  17. covered by this version marker.
  18. The base prompt and bundled skills in this repository are sourced from the
  19. Claude Science product. The public product page is available for context:
  20. https://claude.com/product/claude-science.
  21. ## Repository Layout
  22. | Path | Purpose |
  23. | --- | --- |
  24. | `SYSTEM_PROMPT.md` | Base system prompt and operating rules for the scientific computing agent. |
  25. | `skills/` | Skill packages. Each skill is centered on a `SKILL.md` file, with optional helper code, references, requirements, and provider metadata. |
  26. | `agents/` | Agent profile metadata and prompts for specialized roles such as onboarding, transcript review, bookmarking, and the general science agent. |
  27. | `mcp-servers/` | Bundled MCP servers and widgets used by the agent for external scientific data and chemistry workflows. |
  28. ## Skill Catalog
  29. The `skills/` directory includes workflows for several broad areas:
  30. - Biomolecular modeling and design: `alphafold2`, `boltz`, `chai1`,
  31. `openfold3`, `esmfold2`, `fair-esm2`, `diffdock`, `proteinmpnn`,
  32. `ligandmpnn`, `solublempnn`, `evo2`, and `borzoi`.
  33. - Scientific analysis and reporting: `literature-review`, `pdf-explore`,
  34. `figure-style`, `figure-composer`, `paper-narrative`, and
  35. `indication-dossier`.
  36. - Single-cell and omics workflows: `scgpt` and `scvi-tools`.
  37. - Runtime and environment operations: `remote-compute-modal`,
  38. `remote-compute-ssh`, `compute-env-setup`, `managed-model-endpoints`, and
  39. `using-model-endpoint`.
  40. - Product and customization workflows: `customize`, `skill-creator`,
  41. `product-self-knowledge`, and `self-awareness`.
  42. Most skills start with YAML front matter:
  43. ```yaml
  44. ---
  45. name: example-skill
  46. description: When and why this skill should be loaded.
  47. license: Apache-2.0
  48. ---
  49. ```
  50. The body of `SKILL.md` should give the operational procedure, expected inputs,
  51. failure modes, validation steps, and any tool-specific constraints needed for
  52. reliable execution.
  53. ## Agents
  54. Agent profiles live under `agents/<name>/metadata.yaml`.
  55. - `operon` is the general-purpose scientific computing agent.
  56. - `onboarding` supports first-run user onboarding.
  57. - `reviewer` audits another agent transcript for unsupported claims,
  58. fabrication, or plan deviation.
  59. - `bookmarker` selects transcript spans worth preserving as navigation
  60. breadcrumbs.
  61. These files are configuration assets, not standalone applications. Keep prompt
  62. changes tightly scoped and preserve any eval or benchmark notes near the prompt
  63. they justify.
  64. ## MCP Servers
  65. The `mcp-servers/` directory contains bundled MCP integrations:
  66. - `bio-tools` vendors multiple Python packages for biological and biomedical
  67. data retrieval. `run_server.py` launches a named stdio server from
  68. `mcp-servers/bio-tools/lib/`.
  69. - `ketcher-chemistry` contains a chemistry MCP server and widget assets for
  70. structure editing and chemistry workflows.
  71. The MCP code is runtime infrastructure. When changing it, prefer narrow edits
  72. and verify both the server entry point and the shape of returned tool data.
  73. ## Working With This Repo
  74. Useful inspection commands:
  75. ```bash
  76. # List all skills
  77. find skills -maxdepth 2 -name SKILL.md | sort
  78. # List agent profiles
  79. find agents -maxdepth 2 -name metadata.yaml | sort
  80. # Show launchable bio-tools MCP servers
  81. find mcp-servers/bio-tools/lib -maxdepth 2 -name server.py \
  82. | sed 's#^mcp-servers/bio-tools/lib/##; s#/server.py$##' \
  83. | sort
  84. ```
  85. Before editing a skill or prompt:
  86. 1. Read the relevant `SKILL.md` or `metadata.yaml` completely.
  87. 2. Preserve existing safety, artifact, provenance, and validation rules unless
  88. the change explicitly updates them.
  89. 3. Keep examples executable and aligned with the current runtime surface.
  90. 4. Add or update references next to the skill when a workflow depends on
  91. external documentation or non-obvious operational knowledge.
  92. ## Maintenance Notes
  93. - Use ASCII in new files unless the surrounding file already requires a wider
  94. character set.
  95. - Avoid broad rewrites of prompts or generated MCP bundles without a concrete
  96. regression target.
  97. - Treat bundled third-party services, model weights, and APIs as dependencies:
  98. document licenses, terms, privacy implications, and data movement in the
  99. relevant skill metadata when applicable.
  100. - If a skill ships helper code such as `kernel.py`, keep the markdown
  101. instructions and helper APIs in sync.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.