CtrlK
BlogDocsLog inGet started
Tessl Logo

mlava/scholar-sidekick-mcp

Use the connected scholar-sidekick-mcp MCP server when the user mentions a scholarly identifier (DOI, PMID, PMCID, ISBN, arXiv, ISSN, NASA ADS bibcode, WHO IRIS URL) and wants structured metadata, a formatted citation, a bibliography export file, a retraction check, an open-access check, verification that a claimed citation is real (not fabricated), or a whole-bibliography audit (batch fabrication + retraction check). Requires the MCP server connected — it works anonymously, no API key required; for a zero-install path use the scholar-sidekick-api skill instead.

73

Quality

92%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

Overview
Quality
Evals
Security
Files

Quality

Content

85%

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

A well-structured, highly actionable MCP tool-routing skill with a clear stepped workflow and explicit validation/error-handling checkpoints. Its only real weakness is mild verbosity from repeated warnings and a decorative citation that could be trimmed for token efficiency.

Suggestions

Deduplicate the 'don't run tool names as shell commands' guidance — state it once in Step 0 and reference it rather than repeating near-verbatim in the same section.

Consolidate the 'single identifier per call' notes for checkRetraction/checkOpenAccess/verifyCitation into one place instead of restating across Steps 1 and 3.

Drop the illustrative '(Topaz et al., Lancet 2026)' study citation from the verifyCitation bullet; the fabrication pattern is already described concretely without it.

DimensionReasoningScore

Conciseness

The body is information-dense and mostly assumes Claude's competence, but repeats guidance — 'Do not try to run the tool names in a terminal' / 'do not try to invoke the tool names as shell commands' appears twice in Step 0, 'single identifier per call' recurs across Steps 1 and 3 — and includes a non-essential flourish citation ('Topaz et al., Lancet 2026'), so it could be tightened.

2 / 3

Actionability

Concrete, executable routing guidance throughout: named tools, their parameters ('text', 'id', 'style', 'format'), accepted values (five builtins plus CSL IDs; bib/ris/csl/endnote-xml/...), and an explicit chaining recipe (resolveIdentifier → formatCitation → exportCitation) — copy-paste ready for an MCP-routing skill.

3 / 3

Workflow Clarity

Clear sequenced Steps 0–5 with an explicit validation checkpoint in Step 0 ('If these tools are not in your available toolset, the server is not connected — say so plainly and stop') and error-handling guidance (a failing entry becomes status:'error' without failing the batch), matching the 'clear sequence with explicit validation' anchor.

3 / 3

Progressive Disclosure

No bundle files exist (references/, scripts/, assets/ absent) and none are needed; the skill is a single self-contained file with well-organized sections (When to Use, How to Use Steps 0–5, Guidelines, When NOT to Use), which satisfies the simple-skill allowance for scoring 3 on well-organized structure without external references.

3 / 3

Total

11

/

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.

A strong description: specific capabilities, natural trigger terms spanning the full identifier family, explicit when-guidance, and a well-differentiated niche with a handoff to a sibling skill. The only minor wrinkle is the imperative 'Use the connected ... MCP server' lead rather than a third-person capability verb, but it follows the sanctioned 'Use when...' pattern from the good examples and contains no first/second-person pronoun.

DimensionReasoningScore

Specificity

Lists multiple concrete actions — 'structured metadata, a formatted citation, a bibliography export file, a retraction check, an open-access check, verification that a claimed citation is real (not fabricated), or a whole-bibliography audit' — matching the 'lists multiple specific concrete actions' anchor.

3 / 3

Completeness

Explicitly answers both what (the seven capabilities) and when ('when the user mentions a scholarly identifier ... and wants ...'), with an explicit trigger clause rather than implied timing.

3 / 3

Trigger Term Quality

Covers natural terms users would say across the identifier family ('DOI, PMID, PMCID, ISBN, arXiv, ISSN, NASA ADS bibcode, WHO IRIS URL') plus task phrasings like 'is this citation real (not fabricated)' and 'whole-bibliography audit', giving good coverage of natural keywords.

3 / 3

Distinctiveness Conflict Risk

Occupies a clear scholarly-identifier/citation niche with distinct triggers and even redirects zero-install users to a separate 'scholar-sidekick-api skill', making conflicts with other skills unlikely.

3 / 3

Total

12

/

12

Passed

Validation

81%

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

Validation13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

13

/

16

Passed

Reviewed

Table of Contents