research-lit
Search and analyze research papers, find related work, summarize key ideas. Use when user says "find papers", "related work", "literature review", "what does this paper say", or needs to understand academic papers.
61
71%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Advisory
Suggest reviewing before use
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/skills-codex/research-lit/SKILL.mdQuality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong skill description that clearly communicates specific capabilities (search, analyze, find related work, summarize), provides explicit trigger guidance with natural user phrases, and is well-scoped to the academic research domain. It follows the recommended pattern closely and uses third-person voice throughout.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: 'Search and analyze research papers, find related work, summarize key ideas.' These are distinct, actionable capabilities. | 3 / 3 |
Completeness | Clearly answers both 'what' (search, analyze, find related work, summarize) and 'when' with an explicit 'Use when...' clause listing specific trigger phrases and a general condition. | 3 / 3 |
Trigger Term Quality | Includes natural phrases users would actually say: 'find papers', 'related work', 'literature review', 'what does this paper say', and 'academic papers'. Good coverage of common variations. | 3 / 3 |
Distinctiveness Conflict Risk | Clearly scoped to academic/research paper domain with distinct triggers like 'literature review', 'related work', and 'academic papers'. Unlikely to conflict with general document or web search skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
42%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill is highly actionable with executable code and clear source-specific instructions, but it is severely bloated — the repeated bash resolution blocks alone account for a huge portion of the content and could be factored into a shared helper or referenced file. The workflow is well-sequenced but lacks explicit validation checkpoints between phases. The entire skill is a monolithic document that would benefit enormously from splitting source-specific details into separate files.
Suggestions
Extract the repeated bash $FETCHER resolution pattern into a single shared snippet or helper script, and reference it from each source section instead of duplicating ~15 lines per source.
Move source-specific details (Zotero, Obsidian, Semantic Scholar, DeepXiv, Exa) into separate reference files (e.g., `sources/zotero.md`) and keep only a summary table with links in the main SKILL.md.
Add explicit validation checkpoints: e.g., 'Before Step 2, verify at least 3 papers were found; if fewer, broaden search terms or ask user for guidance.'
Remove explanatory commentary Claude doesn't need (e.g., 'Why use Semantic Scholar?', 'Zotero annotations are gold', descriptions of what each source provides) to cut token usage by ~30-40%.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~400+ lines. Massive bash resolution blocks are repeated nearly identically for each source (arXiv, Semantic Scholar, DeepXiv, Exa, and again for download). Extensive explanations of source detection, de-duplication logic, and override syntax that could be dramatically compressed. Many sections explain things Claude can infer (e.g., 'Why use Semantic Scholar? Many IEEE/ACM journal papers are not on arXiv.'). | 1 / 3 |
Actionability | Provides fully executable bash code blocks for each data source, concrete CLI invocations with flags, specific glob patterns for file discovery, and a clear output table format. The code is copy-paste ready with proper error handling and fallback chains. | 3 / 3 |
Workflow Clarity | Steps are clearly numbered and sequenced (0a through 6), with some validation (PDF size check, de-duplication). However, there are no explicit validation checkpoints between major phases — no 'verify results before proceeding' gates between search, analysis, and synthesis. The error handling is present but implicit (warn and continue), and there's no feedback loop for when synthesis quality is poor or results are insufficient. | 2 / 3 |
Progressive Disclosure | Monolithic wall of text with no content split into separate files. The massive bash blocks for each source, the detailed de-duplication logic, and the source-specific instructions could all be in referenced files. References to `shared-references/integration-contract.md` exist but no bundle files are provided, and the skill itself contains everything inline rather than using progressive disclosure. | 1 / 3 |
Total | 7 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Commit
a425a71
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.