CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

meta-results-risk-of-bias

Generates the "Risk of Bias" results section for a meta-analysis based on assessment tables and statistics. Use when the user wants to draft the risk of bias analysis text from provided data tables.

Quality

68%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./scientific-skills/Academic Writing/meta-results-risk-of-bias/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

50%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 skill's core workflow — a detailed drafting prompt followed by a concrete formatting script call — is genuinely actionable, and script references are real and correctly pointed. However, the body is buried under substantial auto-generated boilerplate, contains duplicated validation sections and a dangling cross-reference, and shows a fabricated expected output for its validation command, which undermines both conciseness and workflow reliability.

Suggestions

Remove the generic boilerplate sections (Key Features, Required Inputs, Validation and Safety Rules, Failure Handling, Deterministic Output Rules, Completion Checklist) and consolidate the two validation sections into one, keeping only the skill-specific workflow (inputs, drafting prompt, format command).

Replace the fabricated "Expected output format" block under Quick Validation with the actual argparse help text of `scripts/format_result.py --help`, or drop it entirely.

Fix the dangling "See `## Workflow` above for related details" reference in Implementation Details (the Workflow section appears later in the file), and add a short example of the final formatted output (text with inserted "(Figure 1)" citation and appended caption) so the expected deliverable is unambiguous.

DimensionReasoningScore

Conciseness

The body is noticeably verbose with several padded boilerplate sections that teach Claude nothing: "Key Features" repeats the frontmatter description verbatim, and generic template blocks ("Required Inputs", "Validation and Safety Rules", "Failure Handling", "Input Validation", "Deterministic Output Rules", "Completion Checklist") state obvious generic discipline rather than skill-specific knowledge. It is above anchor 1 because the core workflow and usage instructions are skill-specific and non-trivial, so not "extensively" explaining known concepts, but multiple sections are pure padding.

2 / 5

Actionability

The core path is concrete and executable: a full drafting prompt template with explicit requirements (">300 words", "analyze each domain (D1-D5)", "Use specific numbers") and a copy-paste command `python scripts/format_result.py --text "<generated_text>" --language "{{language}}"` that matches the actual script's interface. It stops short of anchor 5 because the "Quick Validation" section shows an expected output ("Result file: meta_results_risk_of_bias_result.md ...") that the command `format_result.py --help` does not actually produce, and no example of the final formatted output is given.

4 / 5

Workflow Clarity

The two-step sequence (draft with the role prompt, then format via the script) is clearly listed and a validation shortcut exists, so it is above anchor 2. But validation checkpoints are unreliable or duplicated: the "Expected output format" block for the validation command is fabricated, "Validation Shortcut" and "Quick Validation" repeat the same command, and "Implementation Details" contains a dangling cross-reference ("See `## Workflow` above for related details" — the Workflow section appears later in the document). These are more than minor gaps, matching anchor 3 ("steps listed but validation gaps").

3 / 5

Progressive Disclosure

Bundle structure is sound at one level: `scripts/format_result.py` and `scripts/validate_skill.py` exist and are referenced with correct, clearly signaled paths, with no nested reference chains. However, the body itself is poorly organized into a single ~180-line file with redundant overlapping sections (two validation sections, three checklists, a duplicated description block) that should be consolidated or split, matching anchor 3 ("some structure but could be better organized").

3 / 5

Total

12

/

20

Passed

Description

73%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, well-scoped description with explicit what and when clauses and domain-specific trigger terms. Its main weaknesses are limited action coverage (the two stated actions overlap and downstream capabilities like figure-citation formatting are omitted) and missing common synonyms such as "RoB" or "quality assessment".

DimensionReasoningScore

Specificity

The description names a concrete domain ("Risk of Bias" results section for a meta-analysis) and one-to-two actions ("Generates ... based on assessment tables and statistics", "draft the risk of bias analysis text from provided data tables"), but the two actions largely overlap and it omits other capabilities of the skill (figure citation formatting, language selection, per-domain D1-D5 analysis). This matches anchor 3 ("Names domain and 1-2 concrete actions, but not comprehensive") rather than anchor 4, which requires several distinct specific actions.

3 / 5

Completeness

Both parts are present: a clear "what" ("Generates the 'Risk of Bias' results section for a meta-analysis based on assessment tables and statistics") and an explicit "when" ("Use when the user wants to draft the risk of bias analysis text from provided data tables"). This satisfies anchor 4; it does not reach anchor 5 because the when-clause is a single trigger without the broader, multiple concrete trigger phrases of the 5-anchor example (e.g., no mention of figure formatting or output deliverables as triggers).

4 / 5

Trigger Term Quality

Natural trigger phrases include "risk of bias", "meta-analysis", "results section", "assessment tables", "draft ... analysis text" — terms a user requesting this deliverable would plausibly say. It falls short of anchor 5 because common synonyms and variants like "RoB", "quality assessment", "bias assessment", or "systematic review" are missing.

4 / 5

Distinctiveness Conflict Risk

The description occupies a clear niche (meta-analysis risk-of-bias results drafting from assessment tables) that a user would almost never confuse with a neighboring skill; trigger terms like "risk of bias" and "meta-analysis" are domain-specific with minimal overlap risk.

5 / 5

Total

16

/

20

Passed

Validation

93%

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

CriteriaDescriptionResult

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/Academic Writing/meta-results-risk-of-bias/SKILL.md

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.