CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

drug-photo

Medication photo to personalised PGx dosage card via Claude vision — snap a pill, get genotype-informed guidance

Quality

69%

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

Fix and improve this skill with Tessl

tessl review fix ./skills/drug-photo/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

67%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.

A well-structured, actionable skill body with concrete commands and clear outputs, weakened by repeated descriptions of the same pipeline and implicit rather than explicit validation/fallback checkpoints. No bundle files exist, so all content is inline in a single well-sectioned file.

Suggestions

Consolidate the duplicated pipeline description: 'Why This Exists', 'Core Capabilities', and 'Workflow' each restate photo-identification and the gzip capability — keep one authoritative workflow section and trim the rest.

Add explicit validation/fallback steps to the workflow: what to do when Claude vision cannot identify the drug, when fuzzy matching finds no candidate, or when the genotype file lacks the relevant gene (currently only the INSUFFICIENT DATA output label hints at this).

Move the 51-drug reference table and citations into a references/ file (e.g., references/drugs.md) to slim the SKILL.md overview, and clarify that pharmgx_reporter.py is a dependency from the sibling pharmgx-reporter skill since no scripts/ exist in this bundle.

DimensionReasoningScore

Conciseness

The body is mostly efficient (compact tables, command blocks), but the pill-identification flow is restated three times ("Why This Exists", "Core Capabilities", "Workflow") and "gzip-compressed .txt.gz supported" appears twice, so it could be tightened.

3 / 5

Actionability

Concrete, copy-paste-ready CLI invocations with flags (--input patient.txt.gz --drug Plavix --dose 30mg), a runnable demo command, and an exact output card example. Minor gaps: the vision-identification step is delegated to an external agent, and pharmgx_reporter.py lives in a sibling skill rather than this bundle.

4 / 5

Workflow Clarity

The 5-step Workflow (Photo → Resolve → Genotype → Lookup → Card) is clearly sequenced and failure outcomes are partially handled via the INSUFFICIENT DATA label, but validation checkpoints (e.g., what to do when vision cannot identify the drug or fuzzy matching fails) are implicit rather than explicit — so not a 5.

4 / 5

Progressive Disclosure

Well-organized sections with no nested references and content that mostly fits inline (the 51-drug table is compact and central). Not a 5: the body is over 50 lines and the drug table plus citations could arguably live in a reference file, and there are no bundle files at all despite cross-skill references.

4 / 5

Total

15

/

20

Passed

Description

57%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 distinctive, concrete one-line description that communicates the core photo-to-dosage-card transformation, but it omits any explicit "when to use" trigger guidance and underuses the natural phrases a user would say when needing pill identification. It reads more like a tagline than a complete skill description.

Suggestions

Add an explicit trigger clause, e.g., "Use when the user sends a photo of a pill or medication and asks what it is or how it relates to their genotype."

Include natural trigger synonyms users would actually say — "identify pill", "drug photo", "what is this medication" — rather than only "medication photo" and "snap a pill".

Mention one more concrete capability (e.g., reads real 23andMe genotype data and checks it against CPIC guidelines) to lift specificity from a tagline to comprehensive coverage.

DimensionReasoningScore

Specificity

Names the domain and 1-2 concrete actions ("Medication photo to personalised PGx dosage card via Claude vision", "snap a pill, get genotype-informed guidance") but is not comprehensive — genotype-file lookup, fuzzy drug matching, and CPIC grounding are absent. Not 4: the coverage gaps exceed 'minor'.

3 / 5

Completeness

The "what" is clear (photo in, genotype-informed dosage card out), but there is no "Use when..." clause or equivalent explicit trigger guidance — the "when" is only weakly implied by "snap a pill", which caps completeness at 3.

3 / 5

Trigger Term Quality

Some relevant natural terms are present ("medication photo", "snap a pill", "dosage"), but common variations users would actually say are missing — "identify pill", "drug photo", "what is this medication". Not 4: keyword coverage is moderate rather than good.

3 / 5

Distinctiveness Conflict Risk

A clear niche — genotype-informed dosage cards from medication photos via PGx — with distinct triggers and minimal overlap risk with any generic photo or medication skill.

5 / 5

Total

14

/

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

metadata_field

'metadata' should map string keys to string values

Warning

Total

15

/

16

Passed

Repository

ClawBio/ClawBio

Path

skills/drug-photo/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.