research-lookup

Compile current scholarly evidence for a scientific manuscript or research brief. Use when the user explicitly asks to gather literature, references, background evidence, competing findings, or a manuscript research packet. Uses Parallel Search by default, Parallel Extract for source verification, Parallel Research for explicitly deep/exhaustive work, optional explicit Parallel Chat, and optional Perplexity only when requested or allowed as a failure fallback.

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with a clear, validated multi-step workflow, but it is verbose from repeated backend-selection explanation and monolithic rather than progressively split into reference files. Tightening redundancy and offloading detail into reference docs would raise the two score-2 dimensions.

Suggestions

Consolidate backend selection into the routing table once and have each backend section reference it, removing the repeated 'never selected by automatic routing' / 'preserved as an alternative' restatements.

Move the detailed output-compatibility, failure-handling, and per-backend flag reference material into a reference file (e.g. references/backends.md) linked from a concise overview, so the SKILL.md body stays a lean entry point.

Add an explicit validate-then-retry checkpoint to the batch-mode section (e.g. inspect each query's result envelope and re-run failed queries) to strengthen the one workflow area that currently only isolates failures.

Dimension	Reasoning	Score
Conciseness	The body avoids explaining concepts Claude already knows and is command-dense, but backend-selection logic is repeated across the routing table, the 'Important compatibility behavior' bullets, and five separate backend sections, so it could be tightened rather than earning the lean score-3 anchor.	2 / 3
Actionability	Provides fully executable, copy-paste-ready bash commands with concrete flags, a complete --context-file JSON example, and inspection commands like 'parallel-cli research processors --json', matching the score-3 anchor rather than the pseudocode score-2 case.	3 / 3
Workflow Clarity	The 'Recommended manuscript workflow' is a clearly sequenced 5-step process with an explicit verification step (Parallel Extract source verification), a shortfall-to-refine-search feedback loop, and a 10-item reference-quality checklist, satisfying the score-3 anchor with validation checkpoints.	3 / 3
Progressive Disclosure	Sections are well-organized and the referenced scripts (research_lookup.py, manuscript_packet.py) are real bundle files, but the ~320-line body is largely a monolithic inline document with detailed routing, failure, and output-compatibility content that could live in one-level-deep reference files, fitting the score-2 anchor rather than the well-split score-3 case.	2 / 3
	Total	10 / 12 Passed

Description

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.

The description is strong: third-person, concrete actions, natural trigger terms, and an explicit 'Use when' clause covering both what and when. Its only minor weakness is verbosity in the backend-routing sentence, but no description-rubric dimension maps to conciseness.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions in third person ('Compile current scholarly evidence for a scientific manuscript or research brief', 'gather literature, references, background evidence, competing findings'), matching the score-3 anchor rather than the single-action score-2 example.	3 / 3
Completeness	Explicitly answers both what ('Compile current scholarly evidence for a scientific manuscript or research brief') and when ('Use when the user explicitly asks to gather literature, references, background evidence, competing findings, or a manuscript research packet'), with an explicit 'Use when' clause that clears the score-2 cap.	3 / 3
Trigger Term Quality	Covers natural user phrasing ('literature', 'references', 'background evidence', 'competing findings', 'manuscript research packet') that a user would actually say, satisfying the good-coverage anchor; it is not the score-2 case because common variations are present, not just one keyword.	3 / 3
Distinctiveness Conflict Risk	Occupies a clear scientific-manuscript research niche with distinct triggers, unlikely to fire for unrelated skills; it is above score-2 because the trigger language is specific rather than a generic 'works with documents'.	3 / 3
	Total	12 / 12 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_field	'metadata' should map string keys to string values	Warning

	Total	15 / 16 Passed

Repository: K-Dense-AI/scientific-agent-skills
Path: skills/research-lookup/SKILL.md
Commit: 757b63b

Reviewed: about 12 hours ago

Table of Contents

Discovery Implementation Validation

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.