agentsclimarketplace

Lattice research skill

Skill Dahe-Labs/lattice-research-skill/skills/lattice-research-skill

Router skill for Lattice research workflow. Use when the user asks for Lattice Research Skill, needs help choosing between finding papers, requesting PDFs, scanning full texts, finding research gaps, anti-gap verification, honesty review, or wording review. This skill routes to child skills and keeps shared evidence rules; it does not directly perform retrieval, PDF scanning, data extraction, or gap discovery.From its SKILL.md

Install
npx -y skills add Dahe-Labs/lattice-research-skill --skill lattice-research-skill

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.0 KB, 590 tokens by cl100k_base, as published. Nobody here has run it

Lattice Research Skill

Purpose

Lattice Research Skill is the router for Lattice research workflow. It should not run the research pipeline itself. It selects the right child skill, preserves shared evidence principles, and keeps the workflow modular.

Route Table

User intentUse skillNotes
Find papers, search literature, audit coverage, import selected DOI records into Zotero, create Request PDF list, prepare request_PDF/doi_list.mdlattice-find-papersOwns compact Zotero/request-PDF handoff and zotero_import_manifest.csv; side-product tables are optional.
Analyze project evidence, discover research gaps, or directly verify proposed gapslattice-gapsDefaults to compact gap discovery. Full data/problem/method trees are optional and only used when explicitly requested.
Check whether a research output is honest and evidence-boundlattice-honestyUse for unsupported claims, abstract-as-full-text, correlation-as-causation, and uncertainty marking.
Review proposal/manuscript wording or supervisor commentslattice-wordingUse for Chinese/English academic wording and reusable wording-rule updates.

Shared Rules

  1. Do not make full-text claims from abstracts.
  2. Do not present a missing PDF or missing variable as a proven research gap.
  3. Do not turn broad phrases such as "unified framework", "multi-factor coupling", "data normalization", or "platform/database" into gaps unless they survive verification.
  4. Keep user-visible answers concise and Chinese by default.
  5. Separate facts, inference, suggestions, uncertainty, and missing evidence.
  6. Preserve the Lattice Find Papers handoff contract: runs/<run_id>/request_PDF/ contains doi_list.md and user-downloaded PDFs; find_papers_outputs/tables/zotero_import_manifest.csv records Zotero import/readiness status. Lattice Gaps reads these artifacts and matching Zotero PDFs/full text; it does not manage Zotero imports.
  7. Default to compact outputs: evidence boundary, high-value signals, up to three candidate gaps, Anti-Gap quick check, and 1-3 next actions.

Default Routing

If the user asks to start from a research topic, use lattice-find-papers.

If the user already has PDFs in request_PDF/, Zotero project literature, or other project evidence and asks for gaps, use lattice-gaps.

If the user gives one or more candidate gaps and asks whether they are real, use lattice-gaps in verify-only mode.

If the user asks whether a report is overclaiming, use lattice-honesty.

If the user asks to polish or respond to comments, use lattice-wording.

What ships with it: 3 files

1.0 KB alongside SKILL.md

Keep looking

Skills are one crate of 326,645. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.