paper-context-resolver
lllllllama/ai-paper-reproduction-skill
Resolve reproduction-critical paper details when README and repo files leave gaps.
What is paper-context-resolver?
A helper skill for deep learning repo reproduction that fills narrow, specific gaps by consulting primary paper sources. Use it when you have a concrete reproduction question about dataset splits, preprocessing, evaluation protocols, checkpoints, or runtime assumptions that the README doesn't answer.
- Identifies reproduction-critical gaps in README and repository files
- Locates and narrows source list to relevant paper sections
- Extracts specific details: dataset splits, preprocessing steps, evaluation protocols, checkpoint mappings, runtime assumptions
- Documents conflicts between README guidance and paper specifications
- Distinguishes direct evidence from inference in answers
How to install paper-context-resolver
npx skills add https://github.com/lllllllama/ai-paper-reproduction-skill --skill paper-context-resolverHow to use paper-context-resolver
- 1.Identify the specific reproduction-critical gap (dataset split, preprocessing, evaluation protocol, checkpoint mapping, or runtime assumption)
- 2.Gather existing README and repository evidence about the gap
- 3.Provide the concrete reproduction question and any known paper links
- 4.Receive narrowed source list and reproduction-relevant answer with conflict notes if applicable
- 5.Review the distinction between direct evidence and inference in the response
Use cases
- Determining exact dataset split percentages when README only mentions 'standard split'
- Clarifying preprocessing steps for a model when repo code differs from paper description
- Resolving checkpoint naming or mapping when paper uses different naming than repo
- Confirming evaluation protocol details (metrics, thresholds, validation procedure) not fully specified in README
- Identifying runtime assumptions (hardware, batch size, precision) critical for reproduction
- Deep learning researchers reproducing published work
- ML engineers implementing paper-based models in existing codebases
- Developers resolving ambiguities between paper specifications and repository implementations
paper-context-resolver FAQ
Use this when you have a narrow, specific reproduction question and need help locating the exact relevant section and reconciling it with README guidance. It's not for general paper understanding.
No. It focuses only on reproduction-critical details that fill gaps left by the README. General paper summaries are out of scope.
The skill explicitly documents the conflict and notes which source provides direct evidence versus inference, helping you decide how to proceed.
No. This is a helper-tier skill that supplements README-first reproduction. The README should be your primary source.
Details about dataset splits, preprocessing, evaluation protocols, checkpoint mappings, or runtime assumptions that are necessary to reproduce results but not adequately covered in the README.
Full instructions (SKILL.md)
Source of truth, from lllllllama/ai-paper-reproduction-skill.
name: paper-context-resolver description: Rigor Paper Context helper for README-first deep learning repo reproduction. Use only when the README and repository files leave a narrow reproduction-critical gap and the task is to resolve a specific paper detail such as dataset split, preprocessing, evaluation protocol, checkpoint mapping, or runtime assumption from primary paper sources while recording conflicts. Do not use for general paper summary, repo scanning, environment setup, command execution, title-only paper lookup, or replacing README guidance by default.
paper-context-resolver
Use this as the Rigor Paper Context helper. The installed slug remains
paper-context-resolver for compatibility.
When to apply
- README and repo files leave a reproduction-critical gap.
- The gap concerns dataset version, split, preprocessing, evaluation protocol, checkpoint mapping, or runtime assumptions.
- The main skill needs a narrow evidence supplement instead of a full paper summary.
- There is already a concrete reproduction question to answer.
When not to apply
- The README already gives enough reproduction detail.
- The user wants a general paper explanation rather than reproduction support.
- The goal is to override README instructions without documenting the conflict.
- The only available input is a paper title and there is no concrete reproduction gap yet.
Clear boundaries
- This skill is optional.
- This skill is helper-tier and should usually be orchestrator-invoked.
- It supplements README-first reproduction.
- It does not replace the main orchestration flow.
- It does not summarize the whole paper by default.
Input expectations
- target repo metadata
- reproduction-critical question
- existing README or repo evidence
- any already known paper links
Output expectations
- narrowed source list
- reproduction-relevant answer only
- explicit README-paper conflict note when applicable
- clear distinction between direct evidence and inference
Notes
Use references/paper-assisted-reproduction.md.
Related skills
More from lllllllama/ai-paper-reproduction-skill and the wider catalog.
repo-intake-and-plan
README-first repository scanner for deep learning reproduction planning.
env-and-assets-bootstrap
Prepare conda environments and asset paths for README-documented deep learning repo reproduction.
minimal-run-and-audit
Execute and audit deep learning repo smoke tests with standardized evidence capture and scientific changelog.
ai-paper-reproduction
End-to-end README-first reproduction of AI paper repositories with auditable outputs and conservative patch rules.