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-duvall-trust-and-authority/references/

Source outline

Source: Paul Duvall, Who Gets the Keys? Trust and Authority in Software Factories, AI Native DevCon New York 2026. Recording. All anchors refer to original numbered cues in the transcript. Short quotations below preserve the supplied transcription, including spoken wording.

TopicSource anchorExact short quoteApplication and scope
Upstream judgment00:02:35–00:04:40, cues 37–64“the judgment belongs upstream.”Intent and implementation verification are distinct.
Keep existing checks00:04:44–00:05:49, cues 65–80“That all applies.”The proposed approach retains baseline delivery and security checks.
Tests and reviewers00:05:53–00:07:10, cues 81–99“on its own it's not evidence.”Coverage alone does not show that tests check meaningful behavior.
Containment00:07:20–00:10:09, cues 101–136“the fix is going to be containment.”Untrusted inputs require capability boundaries.
Evidence categories00:10:23–00:13:29, cues 139–179“the evidence decides what ships.”Evidence does not itself determine who has authority.
Agent authority00:13:36–00:14:59, cues 181–198“authorship, what the agents are able to do is not authority.”The speaker separates proposing work from production keys.
Reported incidents00:15:46–00:17:55, cues 210–235“the keys were in the wrong place.”Incident accounts illustrate credential placement risks.
Protected pipeline00:18:07–00:20:28, cues 238–267“pinned one and none.”The proposed pipeline uses pinned inputs and short-lived identity.
Drift types00:20:41–00:22:07, cues 270–289“These are three types of drift.”Different drift affects production, pipeline, and spec trust.
Break glass00:22:10–00:22:53, cues 290–300“the honest exception”Emergency human rollback qualifies the no-standing-access framing.
Contained injection scenario00:22:56–00:25:19, cues 301–328“We contained it.”The scenario illustrates rejecting an unapproved dependency.
Starting steps00:26:44–00:28:17, cues 348–369“no evidence no no deployment.”Machine-checkable controls should trace to accepted intent.

Source limitations

  • This is a proposed architecture and set of reported incidents, not a verified deployment or complete security proof.
  • The transcript does not provide executable policies, repository URLs from the QR code, attestation schemas, or exhaustive threat coverage.
  • Mutation scores, signatures, and independent model judgments provide bounded evidence, not guarantees of semantic correctness.
  • Incident details, OWASP rankings, and third-party throughput claims are talk-time assertions; verify current official sources for operational guidance.
  • Promotional interlude cues 199–209 are omitted from the bundled transcript; the remaining cue numbers retain their source positions.
  • This is a spoken transcript; unseen slides, linked repositories, and third-party research were not independently inspected. No outcome beyond the described talk is established.

skills

talk-duvall-trust-and-authority

tile.json