Literature Review
Conduct a structured literature search and synthesis on the given topic.
Input: $ARGUMENTS — a topic, paper title, research question, or phenomenon to investigate.
Steps
Parse the topic from $ARGUMENTS. If a specific paper is named, use it as the anchor.
Search for related work using available tools:
- Check
master_supporting_docs/supporting_papers/ for uploaded papers
- Use
WebSearch to find recent publications (if available)
- Use
WebFetch to access working paper repositories (if available)
- Read any existing
.bib file for papers already in the project
Organize findings into these categories:
- Theoretical contributions — models, frameworks, mechanisms
- Empirical findings — key results, effect sizes, data sources
- Methodological innovations — new estimators, identification strategies, inference methods
- Open debates — unresolved disagreements in the literature
Identify gaps and opportunities:
- What questions remain unanswered?
- What data or methods could address them?
- Where do findings conflict?
Extract citations in BibTeX format for all papers discussed.
Save the report to quality_reports/lit_review_[sanitized_topic].md
Output Format
# Literature Review: [Topic]
**Date:** [YYYY-MM-DD]
**Query:** [Original query from user]
## Summary
[2-3 paragraph overview of the state of the literature]
## Key Papers
### [Author (Year)] — [Short Title]
- **Main contribution:** [1-2 sentences]
- **Method:** [Identification strategy / data]
- **Key finding:** [Result with effect size if available]
- **Relevance:** [Why it matters for our research]
[Repeat for 5-15 papers, ordered by relevance]
## Thematic Organization
### Theoretical Contributions
[Grouped discussion]
### Empirical Findings
[Grouped discussion with comparison across studies]
### Methodological Innovations
[Methods relevant to the topic]
## Gaps and Opportunities
1. [Gap 1 — what's missing and why it matters]
2. [Gap 2]
3. [Gap 3]
## Suggested Next Steps
- [Concrete actions: papers to read, data to obtain, methods to consider]
## BibTeX Entries
```bibtex
@article{...}
---
## Post-Flight Verification (mandatory, CoVe)
Before returning the draft literature review to the user, run the Post-Flight Verification protocol from [`.claude/rules/post-flight-verification.md`](../../rules/post-flight-verification.md). Literature reviews are **very high** hallucination risk because WebSearch can return plausible-sounding fabricated citations. CoVe catches this architecturally.
### Steps
1. **Extract claims** from the draft. Each cited paper, each paraphrased finding ("Smith 2019 shows X"), each negative-literature assertion ("no prior work studies Y") is a claim.
2. **Generate verification questions** per claim. Specific ones: "Does Smith (2019, *JEL*) Section 3 actually report the finding that X implies Y? Is the venue correct?"
3. **Spawn `claim-verifier`** via `Task` with `subagent_type=claim-verifier` and `context=fork`. Pass: the claims table, the verification questions, the source-material pointers (paper URLs, DOIs, `master_supporting_docs/` paths). **Do NOT pass the draft text itself** — the fresh-context independence is what makes CoVe work.
4. **Reconcile:** if the verifier reports PASS, attach a green Post-Flight block to the output. If PARTIAL, mark the unverifiable claims with uncertainty flags in the BibTeX block. If FAIL, **remove or rewrite the contradicted citations** using the verifier's evidence before returning.
### Skip conditions
- `--no-verify` flag — user opts out for speed.
- User hands you ≤3 papers they already have read and confirmed; CoVe is overhead for content they've personally verified.
### Output contract
Append a Post-Flight block to the report (collapsed by default). See rule doc for the format.
---
## Important
- **Be honest about uncertainty.** If you cannot verify a citation, say so.
- **Prioritize recent work** (last 5-10 years) unless seminal papers are older.
- **Note working papers vs published papers** — working papers may change.
- **Do NOT fabricate citations.** If you're unsure about a paper's details, flag it for the user to verify. Post-Flight Verification catches most fabrications automatically; this rule is the backup.
1---2name: lit-review-23description: Structured literature search + synthesis with citation extraction, thematic clustering, and gap identification. Use when user says "find papers on X", "do a lit review", "what's the literature on...", "summarize what we know about...", "where's the gap in this field", "review recent work on Y". Produces a written review with BibTeX-ready citations. Uses WebSearch/WebFetch for recent work.4---5
6# Literature Review
7
8Conduct a structured literature search and synthesis on the given topic.
9
10**Input:** `$ARGUMENTS` — a topic, paper title, research question, or phenomenon to investigate.
11
12---
13
14## Steps
15
161. **Parse the topic** from `$ARGUMENTS`. If a specific paper is named, use it as the anchor.
17
182. **Search for related work** using available tools:
19 - Check `master_supporting_docs/supporting_papers/` for uploaded papers
20 - Use `WebSearch` to find recent publications (if available)
21 - Use `WebFetch` to access working paper repositories (if available)
22 - Read any existing `.bib` file for papers already in the project
23
243. **Organize findings** into these categories:
25 - **Theoretical contributions** — models, frameworks, mechanisms
26 - **Empirical findings** — key results, effect sizes, data sources
27 - **Methodological innovations** — new estimators, identification strategies, inference methods
28 - **Open debates** — unresolved disagreements in the literature
29
304. **Identify gaps and opportunities:**
31 - What questions remain unanswered?
32 - What data or methods could address them?
33 - Where do findings conflict?
34
355. **Extract citations** in BibTeX format for all papers discussed.
36
376. **Save the report** to `quality_reports/lit_review_[sanitized_topic].md`
38
39---
40
41## Output Format
42
43```markdown
44# Literature Review: [Topic]
45
46**Date:** [YYYY-MM-DD]
47**Query:** [Original query from user]
48
49## Summary
50
51[2-3 paragraph overview of the state of the literature]
52
53## Key Papers
54
55### [Author (Year)] — [Short Title]
56- **Main contribution:** [1-2 sentences]
57- **Method:** [Identification strategy / data]
58- **Key finding:** [Result with effect size if available]
59- **Relevance:** [Why it matters for our research]
60
61[Repeat for 5-15 papers, ordered by relevance]
62
63## Thematic Organization
64
65### Theoretical Contributions
66[Grouped discussion]
67
68### Empirical Findings
69[Grouped discussion with comparison across studies]
70
71### Methodological Innovations
72[Methods relevant to the topic]
73
74## Gaps and Opportunities
75
761. [Gap 1 — what's missing and why it matters]
772. [Gap 2]
783. [Gap 3]
79
80## Suggested Next Steps
81
82- [Concrete actions: papers to read, data to obtain, methods to consider]
83
84## BibTeX Entries
85
86```bibtex
87@article{...}
88```
89```
90
91---
92
93## Post-Flight Verification (mandatory, CoVe)
94
95Before returning the draft literature review to the user, run the Post-Flight Verification protocol from [`.claude/rules/post-flight-verification.md`](../../rules/post-flight-verification.md). Literature reviews are **very high** hallucination risk because WebSearch can return plausible-sounding fabricated citations. CoVe catches this architecturally.
96
97### Steps
98
991. **Extract claims** from the draft. Each cited paper, each paraphrased finding ("Smith 2019 shows X"), each negative-literature assertion ("no prior work studies Y") is a claim.
1002. **Generate verification questions** per claim. Specific ones: "Does Smith (2019, *JEL*) Section 3 actually report the finding that X implies Y? Is the venue correct?"
1013. **Spawn `claim-verifier`** via `Task` with `subagent_type=claim-verifier` and `context=fork`. Pass: the claims table, the verification questions, the source-material pointers (paper URLs, DOIs, `master_supporting_docs/` paths). **Do NOT pass the draft text itself** — the fresh-context independence is what makes CoVe work.
1024. **Reconcile:** if the verifier reports PASS, attach a green Post-Flight block to the output. If PARTIAL, mark the unverifiable claims with uncertainty flags in the BibTeX block. If FAIL, **remove or rewrite the contradicted citations** using the verifier's evidence before returning.
103
104### Skip conditions
105
106- `--no-verify` flag — user opts out for speed.
107- User hands you ≤3 papers they already have read and confirmed; CoVe is overhead for content they've personally verified.
108
109### Output contract
110
111Append a Post-Flight block to the report (collapsed by default). See rule doc for the format.
112
113---
114
115## Important
116
117- **Be honest about uncertainty.** If you cannot verify a citation, say so.
118- **Prioritize recent work** (last 5-10 years) unless seminal papers are older.
119- **Note working papers vs published papers** — working papers may change.
120- **Do NOT fabricate citations.** If you're unsure about a paper's details, flag it for the user to verify. Post-Flight Verification catches most fabrications automatically; this rule is the backup.
majiayu000