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-podjarny-trustworthy-software-factories/references/

Transcript outline

Use the original cue numbers in the transcript. Promotional cues 262–272 are omitted there. This navigation aid adds no independent evidence.

TimeCuesTopic
00:00:51,423–00:02:57,38314–44Factory framing and reported adoption figures
00:03:43,943–00:06:12,46355–95Delegation and skills-to-loops composition
00:06:37,503–00:07:34,303101–118Ticket-to-PR and shared components
00:09:06,063–00:10:10,423142–160Narrow vulnerability-remediation benchmark and workflow
00:10:13,783–00:12:13,423161–195Separate internal remediation and cost reports
00:12:18,263–00:13:39,863196–219Containment versus limits on empowerment
00:13:44,783–00:16:16,023220–261Out-of-scope behavior and verifiers
00:17:17,415–00:18:35,655278–299Loop outputs, telemetry and improvement
00:18:49,935–00:21:08,815303–341Correct behavior before optimization and internal costs
00:21:16,035–00:23:44,435342–382Reusable library, operations and product outcomes
00:23:49,915–00:27:28,275383–442Organizational journey and future aspirations
00:27:48,715–00:33:08,095448–535Partner framing, product demo and dated offers

Exact short quotations

  • “Does my code conform to the skill?” — 00:16:10,463, cue 259.
  • “Stay. Stay in the sandbox to not escalate permissions.” — 00:12:57,423, cue 207.

Evidence limits

  • 26% is a reported vulnerability-remediation success-without-breakage result across 6,000 permutations, not general coding productivity.
  • More than 95% reaching merge is an internal process outcome, not proof of correctness or a controlled comparison.
  • Internal cost figures are workload-specific reports.
  • Model verifiers do not replace enforced tool boundaries.
  • Product demo, credits and proposed GTM workflows are dated or aspirational.

skills

talk-podjarny-trustworthy-software-factories

tile.json