agentsclimarketplace

Finance literature review

Skill azazhang/finance-literature-review-skill/providers/codex/finance-literature-review

Finance-first literature review skill pack for Codex, Cursor, Claude, Gemini, and Copilot.

Install
npx -y skills add azazhang/finance-literature-review-skill --skill finance-literature-review

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

  • 1 stars1 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.

What its author says it does

Copied from the file, not written here

Conduct finance-first literature reviews and related-work synthesis. Use when the user asks for a literature review, related literature section, annotated bibliography, gap map, journal-targeted reading list, finance or macro-finance paper scan, corporate finance or asset-pricing synthesis, banking or disclosure literature summary, or adjacent economics / accounting review. Prioritize top finance journals first, then adjacent top economics and accounting journals, then clearly labeled working papers from NBER, CEPR, SSRN, and RePEc.

SKILL.md

5.1 KB, as published. Nobody here has run it

Finance Literature Review

What This Skill Does

This skill turns a broad literature request into a disciplined finance research workflow:

  1. scope the question,
  2. define the venue ladder,
  3. collect and verify sources,
  4. extract finance-relevant study attributes,
  5. synthesize by mechanism and design,
  6. produce a reading list, matrix, and gap statement.

See:

When To Use

Use this skill when the user wants:

  • a literature review or related literature section,
  • an annotated bibliography,
  • a finance reading list,
  • a working-paper scan,
  • a referee-style map of the literature,
  • a thesis / dissertation literature chapter start,
  • a topic scan across finance, economics, and accounting.

Hard Rules

  • Never fabricate citations or metadata.
  • Always separate published papers from working papers.
  • Prefer top finance journals first.
  • Only widen to adjacent economics or accounting journals when they are genuinely relevant to the question.
  • For empirical papers, capture sample, period, unit of observation, identification strategy, key result, mechanism, and main limitation.
  • If only titles or abstracts are available, say so directly.

Default Source Order

  1. Published top finance journals.
  2. Adjacent top economics journals.
  3. Top accounting journals when the topic overlaps disclosure, reporting, auditing, governance, or information environment.
  4. Working-paper outlets: NBER, CEPR, SSRN, RePEc / IDEAS.
  5. Broad metadata indices and cross-disciplinary sources.

See references/journal-priority.md.

Workflow

Step 1: Scope The Review

Clarify:

  • topic,
  • subfield,
  • mechanism,
  • geography,
  • sample period,
  • whether the user wants published-only or frontier coverage,
  • final output type.

If the request is vague, ask the smallest number of scoping questions needed to fix the venue ladder and the search terms.

Step 2: Build A Search Grid

Define:

  • core concepts,
  • finance-specific synonyms,
  • adjacent economics / accounting terms,
  • method terms,
  • regulation / event names,
  • dataset names,
  • frontier working-paper outlets.

Step 3: Collect Sources

Prioritize seminal published papers first, then recent papers, then frontier working papers.

Do not let frontier working papers crowd out the canonical published literature in a mature area.

Step 4: Extract A Finance Matrix

Use the matrix template in templates/literature-matrix.md.

For each important paper, capture:

  • citation,
  • venue tier,
  • published vs working paper,
  • research question,
  • sample and period,
  • unit of observation,
  • identification strategy,
  • key finding,
  • mechanism,
  • main limitation,
  • why it matters for the user's question.

Step 5: Synthesize

Prefer synthesis by:

  • mechanism,
  • identification strategy,
  • sample or institutional setting,
  • consistency vs disagreement in results,
  • published evidence vs frontier evidence.

Avoid paper-by-paper summaries unless the user explicitly asks for them.

Step 6: Deliver

Default deliverables:

  • scoped review question,
  • prioritized reading list,
  • literature matrix,
  • synthesis,
  • research gaps,
  • citation block.

Before finishing, run the checklist in references/synthesis-checklist.md.

Finance-Specific Heuristics

  • If the topic is asset pricing, keep theory, empirics, and measurement separate.
  • If the topic is corporate finance, separate mechanism papers from reduced-form identification papers.
  • If the topic is banking or macro-finance, be explicit about policy regime and sample window.
  • If the topic overlaps accounting, flag whether the evidence is about disclosure, reporting quality, auditing, or governance.
  • If a result is still mostly a working-paper literature, say that the evidence base is provisional.

Optional Tooling

  • Use Zotero if a structured citation library is available.
  • Use metadata APIs or official journal pages before relying on scraped summaries.
  • Use PDF extraction only when the text is actually needed for design or mechanism verification.

Recommended Output Mode

Pick one default mode unless the user asks for more than one:

  • reading list,
  • literature matrix,
  • related literature section,
  • gap map.

See references/output-modes.md.

Keep looking

Skills are one crate of 328,083. 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.