econ-skills (zbsaygin)¶
external · status: active · focus: analysis · discipline: economics · started: 2026
Project page: https://github.com/zbsaygin/econ-skills
Licence: MIT
Source: projects/landscape/econ-skills.yml
Positioning¶
A personal Claude Code skill pack for academic economics research workflows: reading papers by citekey, multi-agent literature review with novelty assessment, a Delphi-consensus correctness audit of math and econometrics, Econometrica-style notation cleanup, and LaTeX document/Beamer-slide/figure production. Spans literature discovery through paper-drafting and presentation output.
Distinctive contribution¶
notation-clean is a narrow, unusual skill not found in any other cataloged pack: it audits and simplifies mathematical notation against an explicit "Econometrica publication standard" (Greek-for-parameters vs. Latin-for-variables, a 5-use threshold for introducing new symbols, reserved-symbol collision checks), triggered specifically on notation-focused requests rather than general correctness review. audit-econ pairs this with a Delphi-consensus multi-agent audit of derivations, econometrics, code, and data.
Evaluation scores¶
| Dimension | Score (0–3) | Note |
|---|---|---|
| Lifecycle coverage | 2 | Five stages (literature discovery/synthesis, notation/math revision, drafting, presentation output) but no data-analysis, research-design, or referee-simulation skill. |
| Autonomy level | 1 | audit-econ presents severity-sorted findings for the user to triage and 'manages fix workflow' rather than auto-committing fixes; lit-review stress-tests via an adversarial referee but surfaces the gap for the user. |
| Architectural transparency | 3 | All seven skills are public Markdown with explicit sub-agent and consensus-mechanism descriptions; no closed components. |
| Inputs supported | 2 | Multiple input forms (citekey, pasted PDF, LaTeX document, .bib file) plus a literature-access pathway via lit-review's multi-agent search; no private data-source access. |
| Outputs / reproducibility | 2 | Persists compiled LaTeX documents, Beamer decks, and standalone-PDF figures/tables, but requires per-user placeholder setup (paths, Zotero/Better BibTeX) and has no cross-skill artifact manifest. |
| Internal evaluation | 0 | No test fixtures, benchmark, or reported evaluation for any of the seven skills at scoring date. |
| Openness | 2 | MIT license; functional use requires replacing ~10 user-specific placeholders and a Zotero + Better BibTeX bibliography pipeline, so not zero-setup. |
| Maturity / traction | 1 | 7 stars, 1 fork, single commit on record at scoring date — a young, single-maintainer personal tool. |
| Cross-family policy | 0 | No cross-model-family mechanism described; audit-econ's 'consensus' is among same-family Opus sub-agents, not across model families. |
| Runtime assurance | 2 | audit-econ dispatches independent sub-agents to a Delphi consensus before reporting findings; lit-review stress-tests its own novelty claim with an adversarial referee pass until convergence. |
| Cross-platform portability | 0 | Claude Code only; no stated support for other agent runtimes. |
Scored on 2026-10-01. See the evaluation rubric.
Tags¶
Pipeline stages: literature-discovery literature-synthesis revision-editing paper-drafting dissemination
Architectural features: multi-agent debate-consensus tool-use
Inputs: paper-pdf citekey latex-document bibtex-file
Outputs: paper-summary literature-review audit-report latex-document beamer-slides figures
Knowledge sources: zotero-bibtex-library
Limitations¶
- Very young (1 commit, 7 stars); unproven beyond the maintainer's own use.
- Heavy placeholder setup required (project paths, bibliography path, Zotero + Better BibTeX workflow) before any skill runs correctly.
- No reported evaluation or test fixtures for any skill.
- No license file beyond the README's one-line MIT statement.