CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

grasp

Use when the user signals they want to understand something already in play — a result, a document, a file, work done for them — that they cannot follow yet or nod along to unsure; not for an unrelated question or an account that already shows understanding. The user can then say their understanding serves their purpose, on a map checked with them.

Quality

62%

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 ./katalepsis/skills/grasp/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

42%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 body buries a genuinely detailed behavioral protocol inside a monolithic Lean 4 formalization: the sequence and safeguards are there, but at severe token cost and low direct usability. The one occasion-bound reference is a good progressive-disclosure pattern that the rest of the file does not follow.

Suggestions

Move the Lean formal block to a reference file (e.g., references/contract.md) and keep a concise plain-markdown statement of the protocol's steps and safeguards in SKILL.md, loading the formalism only when needed.

Break the `.present` grounding entry's single wall-of-text paragraph into a short list of round-composition rules, each one line, so the instructions are directly usable without reassembly.

Restate the workflow as an explicit numbered sequence (gather → check → map → present → read answer → close) with the validation checkpoints named at each step, instead of leaving them implicit in FLOW/MORPHISM comments.

DimensionReasoningScore

Conciseness

The body is ~450 lines of Lean 4 formalization plus doc comments to convey a dialogue protocol that could be stated in plain markdown in a fraction of the tokens; the `.present` grounding entry alone is a single unbroken paragraph of several hundred words. It is not level 1 because it does not re-explain concepts Claude already knows, but the formal apparatus is padded relative to the behavioral guidance it delivers.

2 / 5

Actionability

Concrete guidance exists (cite file and line for code targets, read references/round-composition.md on stated occasions, the round-composition rules), but the executable core of the skill is delivered as a formal specification with axioms — pseudocode-like — rather than direct instruction, matching 'some concrete guidance but incomplete'.

3 / 5

Workflow Clarity

The FLOW and MORPHISM sections do give a sequence (gather, check, map, present, fuse, close) with an inherent verification loop, but the steps and their checkpoints are implicit, scattered across doc comments inside the formal block rather than presented as a usable workflow.

3 / 5

Progressive Disclosure

The single reference (references/round-composition.md, verified to exist, one level deep, no further nesting) is unusually well-signaled with explicit occasion-bound load conditions, but ~450 lines of formal contract are inlined in SKILL.md — content that clearly belongs in a separate reference file — matching the level-3 anchor of inline content that should be separate.

3 / 5

Total

11

/

20

Passed

Description

62%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 well-triggered, well-bounded description whose 'when' is stronger than its 'what': activation conditions are explicit and even exclude near-miss cases, but the capability itself is conveyed only through an abstract outcome statement rather than concrete actions.

Suggestions

Add one or two concrete capability statements (e.g., 'builds a map of the target's aspects with the evidence each rests on') so the 'what' is as explicit as the 'when'.

Include common user phrasings such as 'explain this result' or 'walk me through this' among the trigger terms to widen natural keyword coverage.

DimensionReasoningScore

Specificity

The description names the domain and target types ("a result, a document, a file, work done for them") and one mechanism ("on a map checked with them"), but never lists concrete actions or capabilities, matching 'names domain and 1-2 concrete actions' rather than the level-4 anchor's 'several specific actions'.

3 / 5

Completeness

Both parts are present: an explicit, even negatively-scoped 'Use when...' trigger, and a 'what' stated as the outcome ("their understanding serves their purpose, on a map checked with them"). It falls short of 5 because the 'what' is outcome-phrased rather than concretely describing what the skill does.

4 / 5

Trigger Term Quality

"understand", "cannot follow", and the target nouns are natural terms, but common user phrasings for this need ("explain this", "walk me through it", "make sense of") are missing, fitting 'some relevant keywords but missing common variations or synonyms'.

3 / 5

Distinctiveness Conflict Risk

The negative triggers ("not for an unrelated question or an account that already shows understanding") carve a distinct niche of verified understanding of in-context material, with only minor overlap risk against general explanation skills.

4 / 5

Total

14

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

jongwony/epistemic-protocols

Path

katalepsis/skills/grasp/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.