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-poulter-factory-improvement/references/

Transcript outline

Use the original cue numbers in the transcript. Promotional cues 200–210 are omitted there. This navigation aid adds no independent evidence.

TimeCuesTopic
00:00:20,380–00:00:47,8607–15Tradeoffs, decomposition and separate lifecycle
00:02:21,380–00:04:20,78038–65Early prompting and oversized specification experience
00:04:26,420–00:07:12,62066–108Typed artifacts and reusable workflow nodes
00:07:15,620–00:08:45,900109–132Retrospective fix that overfit one incident
00:08:52,700–00:10:08,140133–155Local repeats and broader regression comparisons
00:10:12,660–00:13:06,500156–199Startup delay and fixed sandbox constraints
00:13:52,260–00:15:36,340211–237Agent loop, tools and versioned skills
00:15:41,020–00:17:32,260238–268Checkpoint access and dynamic execution
00:17:36,740–00:19:46,340269–299Reported startup and runtime-cost results
00:19:49,220–00:22:36,100300–343CI tradeoffs and shared execution
00:22:38,780–00:25:04,860344–379Symbol graph and earlier feedback
00:25:07,980–00:26:29,380380–401Evidence-based promotion and downside question

Exact short quotations

  • “It's not helping anything else.” — 00:08:22,540, cue 126.
  • “How can you divide work between agents and humans in a governed way?” — 00:04:20,780, cue 65.

Evidence limits

  • Ten repeats and 50–100 regression items are illustrative counts.
  • Repeated inputs do not guarantee identical stochastic outputs.
  • Runtime architecture and security claims are not a complete isolation proof.
  • Startup, cost, agent-count and edit-feedback figures describe one system.
  • Selective checks depend on graph coverage; the talk reports necessary refactoring.

skills

talk-poulter-factory-improvement

tile.json