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

application-template.mdskills/talk-yegge-ai-powered-factories/references/

Personalized factory operating review

This worksheet applies Yegge's reported failure patterns to one factory. It does not assume the user's system behaves like Wheelhouse.

Map the two layers

Reusable execution definitions: Long-lived roles and workstreams: Personal or project-specific history: External relationships: State needed to rebuild: State unsuitable for sharing:

Identify where definitions and accumulated context are mixed. A proposed export should distinguish reusable material from private records and unresolved conversational assumptions.

Investigate one symptom

Observed symptom: Evidence and time window: Work waiting or resources exhausted: Last relevant rule or context change: Competing explanations: Smallest diagnostic check:

Do not label an agent's output as emotional evidence. Describe the observable behavior, such as a role refusing work after a title change or an instruction being generalized beyond its original scope.

Rule and memory maintenance

ItemOriginIntended scopeFreshnessConflict or supersessionProposed handling

Separate temporary requests from durable policy. Identify retention needs before proposing deletion. Preserve the history needed to understand why a decision changed.

Pacing and taste

Output or decision stream: Recipient capacity: Batching or rate proposal: Budget and operational signals: Creative alternatives to compare: User preference captured: Audience represented: Audience missing:

For a game, distinguish experienced-player balance from first-time experience. For other products, name the corresponding underserved journey.

Bounded recovery

Authorized smallest change: Expected observable improvement: Verification period: Rollback or correction path: Maintenance owner: What remains uncertain:

Record whether the work improves a user outcome or merely increases agent activity. Leave time and capacity for maintaining the factory itself. Treat all measurements and trials as planned until evidence of execution exists.

skills

talk-yegge-ai-powered-factories

tile.json