talk-lopopolo-harness-engineering

Use when the user asks about Ryan Lopopolo's AI Native DevCon talk on harness engineering, steering coding agents with goals, constraints, context, tool scope, eval loops, and verification systems.

Quality

74%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./Plugins/aidevcon/skills/talk-lopopolo-harness-engineering/SKILL.md

Quality

Content

75%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 is a lean, well-structured instruction skill with concrete, actionable grounding rules and an explicit transcript-verification step; its main gaps are mild redundancy with the description and referenced bundle files that are not actually present.

Suggestions

Drop the opening framing sentence and the "Core Concepts" list, or move them into outline.md, to remove redundancy with the description and tighten conciseness.

Add a short worked example of a properly cited quote (with transcript line ID) to make the citation guidance copy-paste concrete.

Ensure the referenced outline.md, quote.md, and transcript.md are actually present in the references/ bundle so the signaled progressive-disclosure structure is real, not dangling.

Dimension	Reasoning	Score
Conciseness	The ~40-line body is lean with no padding or re-explanation of concepts Claude already knows, but the opening framing sentence and the "Core Concepts" list restate topics already in the description, minor redundancy that keeps it just below the fully lean score 5.	4 / 5
Actionability	It gives concrete, executable instructions ("Read outline.md first", "verify against transcript.md", "cite the transcript line IDs", "Mark anything beyond the talk as your own recommendation") with specific files and actions, missing only a worked example such as a sample citation format, fitting score 4 rather than fully copy-paste-ready score 5.	4 / 5
Workflow Clarity	The Grounding Rules form a clear sequence with an explicit verification step ("verify against transcript.md when precision matters") and an unsupported-claim escape hatch ("If the transcript does not support a claim, say that"), but it is a prose list rather than an explicit validate-fix-retry checklist, matching score 4.	4 / 5
Progressive Disclosure	The body is well sectioned and clearly signals one-level-deep references (outline.md for locating, quote.md for excerpts, transcript.md for verification), giving good structure; however the referenced files are absent from the bundle (references/scripts/assets do not exist), a real but organization-level gap that fits score 4 rather than the cleanly split score 5.	4 / 5
	Total	16 / 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.

The description is highly distinctive with a clear, explicit trigger and good keyword coverage, but it describes talk topics rather than crisp skill actions and lacks an explicit capability verb, keeping completeness and specificity in the mid-to-upper range.

Suggestions

Lead the description with an explicit action verb (e.g. "Answers questions about" or "Explains") so the skill's capability is stated, not just its subject matter.

Add common synonyms a user might say (e.g. "agent harness", "AI Native DevCon") to broaden natural trigger coverage.

Tighten the topic list into a couple of concrete capabilities (e.g. "summarize, quote, and apply the talk's concepts") to lift specificity toward a comprehensive action set.

Dimension	Reasoning	Score
Specificity	The description enumerates several concrete talk topics ("goals, constraints, context, tool scope, eval loops, and verification") but frames them as subject matter rather than distinct executable capabilities; the actual action (answer/explain the talk) is only implicit, matching the score-3 anchor of naming the domain with limited concrete actions.	3 / 5
Completeness	It has an explicit and specific "when" ("Use when the user asks about...") plus an implied "what" via the enumerated talk topics, but the what lacks a crisp action verb like "Explains/Answers", so it sits at score 4 rather than the fully explicit score 5.	4 / 5
Trigger Term Quality	It opens with the natural trigger phrase "Use when the user asks about Ryan Lopopolo's AI Native DevCon talk on harness engineering" and includes relevant terms like "steering coding agents" and "eval loops" a user would plausibly say, but omits some synonyms (e.g. "agent harness", DevCon alone), fitting score 4 rather than comprehensive score 5.	4 / 5
Distinctiveness Conflict Risk	Naming a specific speaker (Ryan Lopopolo), a specific event (AI Native DevCon), and a specific talk topic gives it a clear niche with minimal overlap risk, matching the score-5 anchor for distinct triggers.	5 / 5
	Total	16 / 20 Passed

Validation

93%

Warnings & errors only

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

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning

	Total	15 / 16 Passed

Repository: jscraik/Agent-Skills
Path: Plugins/aidevcon/skills/talk-lopopolo-harness-engineering/SKILL.md
Commit: f1f2f21

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

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.