CtrlK
BlogDocsLog inGet started
Tessl Logo
PluginPublic

ainativedev/aidevcon-2026-nyc

AI Native DevCon 2026 New York — practical agent skills grounded in conference talks

Quality

86%

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

outline.mdskills/talk-knox-harness-engineering/references/

Source outline

Read transcript.md for exact evidence. Promotion cues 166–176 are omitted.

TimeCuesTopic
00:00–01:351–24Scope and ticket-to-PR definition
01:38–03:4525–55Shared legibility, process improvement and automation
03:51–05:2556–82Delegation, review and shepherd loops
05:30–07:1783–113Factory work as engineering
07:20–10:43114–165Diagnostic baselines and four health metrics
11:27–13:37177–212Five recurring problem categories
13:41–15:36213–245Agent IT, logs and Linear script example
15:42–18:51246–293Architectural nudges and component health
18:54–19:52294–311Shift repeated learning earlier and scan later
19:59–21:36312–338Mutation testing, test bloat and flakiness
21:42–23:29339–366CI test buckets and budgets
23:33–26:05367–406Milestones, orchestration and recap
26:10–30:12407–465Stryker, shaped work and batching Q&A

Short quotations

  • “These are good diagnostic metrics.” — cue 123, 07:54.
  • “correcting the process that produces the code.” — cue 42, 02:50.
  • “factory is” / “for shaped work.” — cues 443–444, 28:42–28:49, separate fragments.

Limits

Baselines are local; PR volume is not a productivity target. Thousand-line files and architectural metrics are nudges, not universal failure thresholds. Review-cost percentages are experience-based suggestions, not study results. Internal automerge is a reported choice with an explicit internal-factory rationale. The audience's independent-test-agent suggestion is not presented as Knox's established workflow.

skills

talk-knox-harness-engineering

tile.json