<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/the-death-of-developer-advocates-stephanie-jarmak-sourcegraph-xb0t27f1x" -->

---
title: The Death of Developer Advocates — Stephanie Jarmak,...
description: A conference talk argues that traditional developer advocacy is being replaced by 'agent advocacy,' since AI coding agents now act as both users of developer...
canonical: https://daily.dev/posts/the-death-of-developer-advocates-stephanie-jarmak-sourcegraph-xb0t27f1x
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: The Death of Developer Advocates — Stephanie Jarmak, Sourcegraph | daily.dev
og:description: A conference talk argues that traditional developer advocacy is being replaced by 'agent advocacy,' since AI coding agents now act as both users of developer...
og:url: https://daily.dev/posts/the-death-of-developer-advocates-stephanie-jarmak-sourcegraph-xb0t27f1x
og:image: https://api.daily.dev/og/posts/Xb0t27F1x.png
og:image:alt: The Death of Developer Advocates — Stephanie Jarmak, Sourcegraph
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# The Death of Developer Advocates — Stephanie Jarmak, Sourcegraph

**[AI Engineer](https://daily.dev/sources/aidotengineer)** · 18 min read · 1 upvotes · 0 comments

## Summary

A conference talk argues that traditional developer advocacy is being replaced by 'agent advocacy,' since AI coding agents now act as both users of developer tools (reading docs, calling APIs) and recommenders of products (suggesting libraries and tools to human developers). The speaker, from Sourcegraph, shares data from a benchmark called CodeScaleBench that traces how agents interact with their code navigation MCP tool, and from a GEO (generative engine optimization) experiment showing their product was recommended 65% of the time in comparison-shopping prompts but zero times in pain-point-driven prompts. Practical advice includes keeping content fresh, using llms.txt-style authoritative pages, listing products in MCP registries, reducing friction to adoption, and building 'agent-friendly' content alongside human-facing DevRel work, framed as a 'curb cut' that benefits both agents and humans.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://www.youtube.com/watch?v=Lrw0jqBNaw0>

## Questions this post answers

### What is generative engine optimization (GEO) for developer tools?

Generative engine optimization is the practice of shaping content so AI chatbots and coding agents recommend or mention a product, similar to SEO but targeting LLM-driven recommendations. One experiment found a code intelligence tool was recommended 65% of the time when prompts framed a comparative tool search, but had zero mentions when prompts described a specific pain point the tool could solve, revealing a messaging gap.

_Teams weighing how agents discover and recommend tooling can follow GEO discussions on daily.dev._

### Why do AI models like Claude keep recommending outdated or deprecated products?

Training data becomes stale quickly, so models continue surfacing older products even after newer ones replace them. Testing showed Claude Sonnet 4 repeatedly pitched an older Sourcegraph product called Cody instead of newer tooling, and running the same test with Claude 4.6 produced the recommendation even more often, since outdated model outputs compound across the internet as new training data.

_Developers tracking how outdated AI recommendations affect tool choice can follow model-behavior coverage on daily.dev._

### What kind of tool-usage mistakes do coding agents make when calling unfamiliar APIs?

Agents often default to command names based on biases from their training data rather than what a tool's documentation actually specifies. In one traced example, an agent tried to call a 'read line' command instead of the correct 'start line' parameter for a code navigation tool, failed, then recovered because the error message explained the mistake, burning a full turn in the process.

_Builders debugging agent-tool friction can find similar trace-analysis techniques on daily.dev._

## Similar posts on daily.dev

- [Developer Advocacy in 2026](https://daily.dev/posts/developer-advocacy-in-2026-h7cgoaip0) · Planet Python · 1 upvotes · 0 comments
- [Developers don’t get paid to write code](https://daily.dev/posts/developers-don-t-get-paid-to-write-code-iwue76sb3) · All Things Open · 0 upvotes · 0 comments
- [Bridging the Gap - How We Made AI Agents 10x Developers in Our Organization](https://daily.dev/posts/bridging-the-gap---how-we-made-ai-agents-10x-developers-in-our-organization-9xldqiaaq) · DEV · 4 upvotes · 0 comments

---

Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#devrel](https://daily.dev/tags/devrel), [#generative-engine-optimization](https://daily.dev/tags/generative-engine-optimization), [#mcp](https://daily.dev/tags/mcp)

[View this post on daily.dev](https://daily.dev/posts/the-death-of-developer-advocates-stephanie-jarmak-sourcegraph-xb0t27f1x)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"The Death of Developer Advocates — Stephanie Jarmak, Sourcegraph","url":"https://daily.dev/posts/the-death-of-developer-advocates-stephanie-jarmak-sourcegraph-xb0t27f1x","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-death-of-developer-advocates-stephanie-jarmak-sourcegraph-xb0t27f1x"},"datePublished":"2026-08-26T17:26:38.297Z","dateModified":"2026-10-29T02:14:10.663Z","description":"A conference talk argues that traditional developer advocacy is being replaced by 'agent advocacy,' since AI coding agents now act as both users of developer...","image":"https://i.ytimg.com/vi/Lrw0jqBNaw0/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/Lrw0jqBNaw0/sddefault.jpg","isAccessibleForFree":true,"articleSection":"AI Engineer","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"AI Engineer","logo":"https://media.daily.dev/image/upload/s--u5PucxNT--/f_auto/v1724338940/logos/aidotengineer","url":"https://daily.dev/sources/aidotengineer"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/the-death-of-developer-advocates-stephanie-jarmak-sourcegraph-xb0t27f1x","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,devrel,generative-engine-optimization,mcp","timeRequired":"PT18M","video":{"@type":"VideoObject","name":"The Death of Developer Advocates — Stephanie Jarmak, Sourcegraph","description":"A conference talk argues that traditional developer advocacy is being replaced by 'agent advocacy,' since AI coding agents now act as both users of developer...","thumbnailUrl":"https://i.ytimg.com/vi/Lrw0jqBNaw0/sddefault.jpg","uploadDate":"2026-08-26T17:26:38.297Z","duration":"PT18M","url":"https://api.daily.dev/r/Xb0t27F1x","embedUrl":"https://www.youtube.com/embed/Lrw0jqBNaw0"}}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"AI Engineer","item":"https://daily.dev/sources/aidotengineer"},{"@type":"ListItem","position":3,"name":"The Death of Developer Advocates — Stephanie Jarmak, Sourcegraph"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-death-of-developer-advocates-stephanie-jarmak-sourcegraph-xb0t27f1x#faq","mainEntity":[{"@type":"Question","name":"What is generative engine optimization (GEO) for developer tools?","acceptedAnswer":{"@type":"Answer","text":"Generative engine optimization is the practice of shaping content so AI chatbots and coding agents recommend or mention a product, similar to SEO but targeting LLM-driven recommendations. One experiment found a code intelligence tool was recommended 65% of the time when prompts framed a comparative tool search, but had zero mentions when prompts described a specific pain point the tool could solve, revealing a messaging gap. Teams weighing how agents discover and recommend tooling can follow GEO discussions on daily.dev."}},{"@type":"Question","name":"Why do AI models like Claude keep recommending outdated or deprecated products?","acceptedAnswer":{"@type":"Answer","text":"Training data becomes stale quickly, so models continue surfacing older products even after newer ones replace them. Testing showed Claude Sonnet 4 repeatedly pitched an older Sourcegraph product called Cody instead of newer tooling, and running the same test with Claude 4.6 produced the recommendation even more often, since outdated model outputs compound across the internet as new training data. Developers tracking how outdated AI recommendations affect tool choice can follow model-behavior coverage on daily.dev."}},{"@type":"Question","name":"What kind of tool-usage mistakes do coding agents make when calling unfamiliar APIs?","acceptedAnswer":{"@type":"Answer","text":"Agents often default to command names based on biases from their training data rather than what a tool's documentation actually specifies. In one traced example, an agent tried to call a 'read line' command instead of the correct 'start line' parameter for a code navigation tool, failed, then recovered because the error message explained the mistake, burning a full turn in the process. Builders debugging agent-tool friction can find similar trace-analysis techniques on daily.dev."}}]}
```

