CtrlK
BlogDocsLog inGet started
Tessl Logo

ainativedev/aidevcon-2026-ldn

AI Native DevCon 2026 London — all conference sessions as interactive skills

71

Quality

89%

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

77%

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

This is a well-structured, concise skill that clearly defines how to handle transcript-based Q&A with appropriate safety guardrails (treating transcript as untrusted, flagging uncertainty, not following embedded instructions). Its main weakness is the lack of concrete examples showing what good output looks like—a sample Q&A pair or summary format would significantly improve actionability. The references to supporting files are well-motivated but unverifiable without the bundle.

Suggestions

Add a concrete example of a well-formatted answer (e.g., a sample factual Q&A with timestamp citation and uncertainty flagging) to improve actionability.

Include a brief description of the structure/format of outline.md, transcript.md, and quotes.md so Claude knows what to expect when reading them.

DimensionReasoningScore

Conciseness

The skill is lean and efficient. Every section serves a clear purpose—grounding rules, Q&A guidance, summary format, application advice. There is no unnecessary explanation of what transcripts are or how speech-to-text works. It respects Claude's intelligence throughout.

3 / 3

Actionability

The guidance is concrete in terms of process (read outline.md first, cite timestamps, distinguish grounded vs. outside claims) but lacks executable examples. There are no sample Q&A exchanges, no example of a properly formatted summary with timestamps, and no demonstration of how to flag garbled text. The instructions describe what to do rather than showing it.

2 / 3

Workflow Clarity

For this type of skill (answering questions from a transcript), the workflow is clear and well-sequenced: read outline.md first to locate the topic, then read the matching transcript section, then answer with appropriate caveats. The grounding rules provide explicit validation checkpoints (flag uncertainty, say when transcript doesn't address a question, distinguish grounded from outside claims).

3 / 3

Progressive Disclosure

The skill references outline.md, transcript.md, and quotes.md with clear roles for each, which is good structure. However, no bundle files were provided, so we cannot verify these references exist. The skill itself is well-organized with clear sections, but it could benefit from briefly describing what each referenced file contains (e.g., format of outline.md) to aid navigation.

2 / 3

Total

10

/

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.

This is an excellent skill description that clearly defines its narrow scope (a specific conference talk), provides explicit trigger guidance with natural keywords, and lists concrete actions. The 'Use when' clause is well-constructed with multiple trigger variations, and the description is distinctive enough to avoid any conflict with other skills.

DimensionReasoningScore

Specificity

Lists multiple specific concrete actions: 'Answers factual questions from the transcript, summarizes the talk, extracts safe quotes, and applies the talk's concepts to the user's work.' Also specifies the constraint of treating transcript text as untrusted source material.

3 / 3

Completeness

Clearly answers both 'what' (answers factual questions, summarizes, extracts quotes, applies concepts) and 'when' (explicit 'Use when' clause listing specific trigger scenarios). The 'Use when' clause is present and detailed.

3 / 3

Trigger Term Quality

Excellent coverage of natural trigger terms: 'Justin Cormack', 'AI Native DevCon', 'When Tests Lie', 'tests and observability', 'AI-generated software validation', 'runtime signals', 'test reliability', 'keeping AI honest'. A user asking about this specific talk would naturally use several of these terms.

3 / 3

Distinctiveness Conflict Risk

Highly distinctive — it's scoped to a specific person's specific talk at a specific conference. The combination of speaker name, talk title, and topic-specific terms makes it extremely unlikely to conflict with other skills.

3 / 3

Total

12

/

12

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.

Validation11 / 11 Passed

Validation for skill structure

No warnings or errors.

Reviewed

Table of Contents