CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

talk-kushwaha-benchmarking-agent-era

Use when the user asks about Amit Kushwaha's AI Native DevCon talk on benchmarking agent-era systems, measuring performance beyond single LLM calls, inference, workflow complexity, tool use, and real-world workloads.

Quality

82%

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

85%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-organized, actionable reference-skill body with a clear grounding workflow and explicit verification and error-handling rules. The one significant weakness is that the three files the entire workflow depends on are absent from the bundle, leaving the progressive-disclosure structure unresolvable as shipped.

Suggestions

Ship `outline.md`, `quote.md`, and `transcript.md` in the skill's `references/` directory so the grounding rules' file targets actually resolve.

If the bundle files were intentionally omitted from this review copy, note their expected locations (e.g., `references/outline.md`) in the body so paths are unambiguous.

Consider trimming or merging the "Core Concepts" list into the opening paragraph, since it duplicates the topic enumeration already stated there.

DimensionReasoningScore

Conciseness

The body is lean and assumes Claude's competence: rules like "Read `outline.md` first to locate the relevant section or concept" and "Attribute claims to Amit Kushwaha; if a line is from the host or an audience member, say so" carry real informational weight. The only trimmable part is the "Core Concepts" bullet list, which largely restates the topic enumeration already given in the opening paragraph, keeping it just short of the every-token-earns-its-place anchor 5.

4 / 5

Actionability

For an instruction-only skill the guidance is fully actionable: it names exact files and their roles ("Read `outline.md` first", "Use `quote.md` for short supporting excerpts, then verify against `transcript.md`"), requires citing "transcript line IDs when available", and covers the three common task modes (factual Q&A, applying the talk, comparing with other talks) each with concrete steps. Per the rubric's scoring note, absence of code is not penalized when guidance is this specific.

5 / 5

Workflow Clarity

The grounding rules form a clearly sequenced checklist: outline first for location, quotes for excerpts, transcript verification "when precision matters", correct attribution, and an explicit failure-handling rule ("If the transcript does not support a claim, say that the talk does not address it"). This is a simple non-destructive skill whose verification and error-recovery steps are explicit, matching anchor 5.

5 / 5

Progressive Disclosure

The in-body structure is good — references are one level deep, clearly signaled with backticks, and each given a role — but the actual bundle contains no `references/`, `scripts/`, or `assets/` directories, so `outline.md`, `quote.md`, and `transcript.md` do not exist alongside the skill. Scoring against the actual bundle structure per the rubric guideline, the primary navigation targets are unresolvable, which breaks the discovery path the skill depends on.

3 / 5

Total

17

/

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, niche description with an explicit and specific trigger clause and excellent distinctiveness. Its main weakness is that it enumerates the talk's topics but never states the skill's concrete capabilities, leaving the 'what' implied rather than explicit.

Suggestions

Add an explicit capability statement before the trigger clause, e.g., "Answers questions, summarizes, and applies the concepts from Amit Kushwaha's AI Native DevCon talk on benchmarking agent-era systems..."

Include a few natural synonyms in the trigger, such as "agent benchmarks", "evals", or "measuring agent performance", to broaden keyword coverage.

DimensionReasoningScore

Specificity

The description names the domain precisely ("benchmarking agent-era systems") and enumerates specific subject areas ("measuring performance beyond single LLM calls, inference, workflow complexity, tool use, and real-world workloads"), but states no concrete actions the skill performs (e.g., answering questions, summarizing, or applying the talk's concepts). It sits above anchor 2, where the domain is named but nothing else is specific, and below anchor 4, which requires several listed capability actions.

3 / 5

Completeness

The 'when' is explicit and specific ("Use when the user asks about Amit Kushwaha's AI Native DevCon talk..."), and a topical 'what' is conveyed through the enumerated subjects ("measuring performance beyond single LLM calls, inference, workflow complexity..."). However, the 'what' never states what the skill actually does, so it falls short of anchor 5's requirement that both what and when are clearly and explicitly answered.

4 / 5

Trigger Term Quality

Phrases like "Amit Kushwaha's AI Native DevCon talk", "benchmarking agent-era systems", "tool use", and "real-world workloads" are exactly what a user asking about this talk would say. Coverage is good but misses common synonyms a user might naturally use such as "evals", "agent benchmarks", or "performance testing", so it does not reach the comprehensive anchor 5.

4 / 5

Distinctiveness Conflict Risk

Anchoring the trigger to a named speaker ("Amit Kushwaha") and a named event ("AI Native DevCon") gives this skill a clear niche with distinct triggers and minimal conflict risk against other benchmarking or performance skills.

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

metadata_version

'metadata.version' is missing

Warning

Total

15

/

16

Passed

Repository

jscraik/Agent-Skills

Path

Plugins/aidevcon/skills/talk-kushwaha-benchmarking-agent-era/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.