capwitf/My-MathModeling-skills / math-literature
math-literature
VerifiedUse when 高教社杯/CUMCM mathematical modeling papers need lightweight literature search, citation verification, method/background references, standards or policy sources, Chinese-English source routing, DOI/BibTeX checks, or claim-to-citation mapping without fabricated sources.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 3.1 KB |
| agents/openai.yaml | config | 306 B |
| references/citation-evidence-contract.md | markdown | 2.4 KB |
| references/contest-scope.md | markdown | 2.3 KB |
| references/search-workflow.md | markdown | 3.1 KB |
| references/source-tiers.md | markdown | 1.9 KB |
references/citation-evidence-contract.md
View on GitHub# Citation Evidence Contract
## Reference Registry
Use `reference_registry.csv` for important sources that may enter the paper. In short contest workflows, ordinary background sources can be recorded in `reference_notes.md` instead.
```csv
reference_id,title,authors,year,venue,doi,url,source_tier,verified_from,supports,limits,status
```
`status` values:
- `candidate`: discovered but not verified.
- `verified`: metadata and relevance checked.
- `paper_ready`: mapped to a specific paper claim or paragraph.
- `blocked`: metadata, access, or relevance is insufficient.
- `rejected`: checked and not suitable.
## Claim Citation Map
Use `claim_citation_map.csv` when citations support paper text.
```csv
map_id,claim_id,location,reference_id,support_type,exact_supported_point,boundary,status
```
`support_type` examples:
- `background`
- `method_precedent`
- `parameter_source`
- `baseline`
- `dataset`
- `standard_or_policy`
- `limitation`
## Paper-Ready Citation Rules
A citation is paper-ready only when:
- the source has a verified registry row;
- the cited statement is narrower than or equal to the source support;
- the paper location is known;
- the source is the primary source when primary evidence is needed;
- the boundary explains what the source does not prove.
For 高教社杯/CUMCM-style papers, apply this strictly to method, parameter, standard, policy, dataset, and comparison claims. Do not require a row-level map for every generic background sentence unless the user or contest asks for a formal source audit.
## Related Work Structure
Write related work by topic:
1. Problem class and application context.
2. Main modeling routes used in prior work.
3. What each route handles well.
4. What remains missing for the active contest problem.
5. How the present model responds to that gap.
Avoid paper-by-paper summaries unless the user explicitly asks for an annotated bibliography.
## Red Flags
Block or downgrade when:
- a DOI, title, venue, year, or author list is guessed;
- a review article is used as if it were the primary source for data;
- a citation is attached to a number that came from the contest data or code output;
- a method is cited but the active model does not actually use that method;
- a Chinese source has only a search snippet and no verified export or page;
- related work claims novelty without naming the baseline or prior limitation.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |