CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

write-conclusion-pack

Turn a finished local model eval into a conclusion pack (conclusions, data tables, charts, run manifest, limitations, raw data) that someone else can check and reproduce.

Quality

87%

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

SKILL.md
Quality
Evals
Security

Quality

Content

93%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.

An exemplary lean, actionable skill body: executable commands, a precisely specified report template, and honest framing of limitations and small-sample caveats. The only meaningful gap is the lack of an explicit verification checkpoint on the finished report, which keeps workflow clarity at 4.

DimensionReasoningScore

Conciseness

The ~40-line body contains no concept explanations and no padding; every line is instructional (e.g., "Charts come out as SVG and, where the system can convert it, PNG (posts need PNG). If only SVG is produced, say so."). It fully assumes Claude's competence, matching the lean anchor 5.

5 / 5

Actionability

Copy-paste-ready bash commands are given for tables and both chart kinds, and each report section has a concrete spec ("three to five one-line judgments, each with the condition it holds under", an explicit manifest field list). Placeholders like <run_dir> are appropriate parameterization rather than gaps, so this fits the fully-executable anchor 5.

5 / 5

Workflow Clarity

The sequence (build data → write report.md → finish) is clear, with one conditional checkpoint ("If only SVG is produced, say so") and the traceability rule ("every claim in it must trace back to a number in the same folder"). However, there is no explicit validate-and-fix loop on the finished report (e.g., verify each conclusion cites a data number), so it matches anchor 4 rather than the explicit-validation anchor 5.

4 / 5

Progressive Disclosure

The skill is under 50 lines, has no bundle files, and needs no external references; content is appropriately split into three well-labeled sections (Build the data, Write report.md, Finish). Per the rubric's scoring notes, this simple structure earns a 5.

5 / 5

Total

19

/

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 specific, well-differentiated description that clearly states what the skill produces and why reproducibility matters. Its main weakness is the absence of any explicit "when to use this" trigger clause, which both caps completeness and leaves some natural trigger phrases (e.g., benchmark write-up, report) uncovered.

Suggestions

Add an explicit trigger clause, e.g. "Use when an eval run has finished and results need to be written up, e.g. after run-model-eval completes" — this would lift completeness from 3 toward 5.

Include one or two natural synonym phrases users might say, such as "benchmark results" or "eval write-up/report", to broaden trigger-term coverage.

Keep the deliverable list as-is; it is the strongest part of the description and should not be trimmed.

DimensionReasoningScore

Specificity

The description enumerates six concrete deliverables ("conclusions, data tables, charts, run manifest, limitations, raw data") plus an explicit purpose ("that someone else can check and reproduce"), giving comprehensive coverage of what the skill produces. It goes beyond the anchor-4 level because nothing material about the output is left unnamed.

5 / 5

Completeness

The "what" is clear and detailed, but there is no "Use when..." clause or equivalent explicit trigger guidance in the description; timing is only weakly implied by "a finished local model eval". Per the rubric guidelines, a missing explicit trigger caps completeness at 3 even with a strong "what".

3 / 5

Trigger Term Quality

Natural phrases like "model eval", "conclusions", "charts", and "reproduce" are terms a user would plausibly say, but common variations such as "benchmark results", "write-up", or "report" are absent. Good coverage with a few natural terms missing matches anchor 4 rather than the synonym-rich anchor 5.

4 / 5

Distinctiveness Conflict Risk

"Turn a finished local model eval into a conclusion pack" carves out a clear niche with distinct triggers; the deliverable list further separates it from generic reporting or post-processing skills. Conflict risk is minimal, matching anchor 5 rather than the minor-overlap anchor 4.

5 / 5

Total

17

/

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

holaboss-ai/holaOS

Path

runtime/api-server/src/embedded-capabilities/local-model-review/skills/write-conclusion-pack/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.