scite-database

Access Scite.ai Smart Citations to classify how a paper is cited (supporting, contrasting, mentioning) and assess scientific claims; use it when you need to evaluate a paper’s reliability or its acceptance in the literature.

Quality

67%

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

Fix and improve this skill with Tessl

tessl review fix ./scientific-skills/Evidence Insight/scite-database/SKILL.md

Quality

Content

46%Weight 40%Scale 1-5

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

The second half of the body is a well-written, concrete guide to a simple single-script skill, but it is buried under ~50 lines of duplicated auto-generated boilerplate that contradicts it (conflicting Python versions, a nonexistent run path, and claims of 4 scripts when only 1 exists). The executable CLI examples are solid, though the sample output labels drift from the script's real output.

Suggestions

Delete the duplicated boilerplate half (the first 'When to Use', 'Key Features', 'Dependencies', 'Example Usage', and 'Implementation Details' sections) and keep only the substantive numbered sections 1-5.

Remove or correct the false bundle claims — 'plus 3 additional script(s)' and 'additional helper scripts under scripts/' — since the bundle contains only scripts/scite_client.py, and fix the broken 'cd 20260316/scientific-skills/...' path in the usage example.

Reconcile the two conflicting dependency blocks into one (Python 3.9+ and requests 2.x per the actual script) and align the example text output with the script's real labels ('support:', 'refute:', 'mention:') or fix the script to match the documented output.

Dimension	Reasoning	Score
Conciseness	The body duplicates nearly every section — two 'When to Use', two 'Key Features', two 'Dependencies', two 'Example Usage', two 'Implementation Details' — with the first half being auto-generated boilerplate that restates the description verbatim ("Use this skill when you need access scite.ai smart citations to classify...") and generic filler like "Output discipline: keep results reproducible". Not a 1 because it never explains concepts Claude already knows; clearly a 2 with several padded/duplicated sections that should be deleted.	2 / 5
Actionability	Concrete executable commands against a real script: `python scripts/scite_client.py "10.1038/nature12345"`, `--format json`, and `--help`, with example text and JSON outputs. Minor gaps: the example cd path '20260316/scientific-skills/...' does not exist in the bundle, and the shown text output ('--- Scite Analysis for... / Supporting: 12') does not exactly match the script's actual output labels ('--- Scite analysis report:', 'support:', 'refute:').	4 / 5
Workflow Clarity	The core flow (run the script with a DOI, choose text or JSON output) is unambiguous for this simple read-only skill, but the duplicated sections create real ambiguity: two conflicting dependency specs ("Python: 3.10+" vs "Python: 3.9+"; "not explicitly version-pinned" vs "requests: 2.x"), two different 'when to use' lists, and a broken cd path in the run example. Not a 4 because these contradictions leave the reader with competing instructions rather than minor gaps.	3 / 5
Progressive Disclosure	The body claims "3 additional script(s)" and "additional helper scripts under scripts/", but the actual bundle contains only scripts/scite_client.py — a false reference that misrepresents the bundle structure. Combined with the fully inlined duplicated boilerplate sections, navigation signals point at content that does not exist. Not a 3 because the structure is not merely improvable; it actively misleads.	2 / 5
	Total	11 / 20 Passed

Description

75%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: it states concrete capabilities in a distinct niche, uses third-person voice, and includes an explicit 'use it when' trigger clause covering reliability and literature-acceptance evaluation. The main gaps are the missing 'DOI' trigger keyword and slightly generic reliability-assessment phrasing.

Dimension	Reasoning	Score
Specificity	"classify how a paper is cited (supporting, contrasting, mentioning) and assess scientific claims" names several concrete actions in a specific domain, but coverage has minor gaps — no mention of the DOI input, tally counts, or output formats. Not a 5 because the action list is not comprehensive; not a 3 because it goes beyond 1-2 actions.	4 / 5
Completeness	The 'what' is clear ("Access Scite.ai Smart Citations to classify how a paper is cited... and assess scientific claims") and the 'when' is explicit ("use it when you need to evaluate a paper's reliability or its acceptance in the literature"). Not a 5 because the when-clause could name more concrete triggers (e.g., literature reviews, fact-checking a claim, DOI lookup).	4 / 5
Trigger Term Quality	Good natural-term coverage: "how a paper is cited", "paper's reliability", "acceptance in the literature", "scientific claims", "supporting, contrasting, mentioning" — phrases a user would naturally say. A few terms are missing, most notably "DOI" (the actual invocation input) and "citation analysis".	4 / 5
Distinctiveness Conflict Risk	"Scite.ai Smart Citations" is a clearly branded niche with distinct triggers, giving minimal conflict risk. Not a 5 because the broader phrases "assess scientific claims" and "evaluate a paper's reliability" could overlap with other literature-evaluation or evidence-assessment skills in a collection.	4 / 5
	Total	16 / 20 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
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: aipoch/medical-research-skills
Path: scientific-skills/Evidence Insight/scite-database/SKILL.md
Commit: 686e09d

Reviewed: 13 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.