# How to export a coding agent session — and make it worth reading · Stacktree

Source: https://stacktr.ee/blog/export-coding-agent-session

[Skip to content](#main) [Stacktree](/)[Developers](/developers)[Agents](/agents)[Docs](/docs)[Use cases](/use-cases)[Pricing](/pricing)[Blog](/blog)[Dashboard](https://app.stacktr.ee)[Sign in →](https://app.stacktr.ee)      By [ Steve Smith ](/about) · Founder, Stacktree  ·  Last updated August 24, 2026          blog · hands-on · receipts included

#  Your best agent session is a 4,000-line markdown file.

   YC now asks applicants to attach a coding agent session they're proud of, and every serious agent tool can export one. What none of them solve: the export is a wall of raw markdown, and the people you want to show it to — a cofounder, a reviewer, the internet — will not scroll a wall. Getting the transcript out, and making it readable, are two different jobs.

    [ Get started free ](https://app.stacktr.ee/?join=1&from=seo_blog_export_coding_agent_session)   No card · 3 pages free · about a minute

##  How do you export and share a coding agent session?

Export it from the tool — `/export` in Claude Code, the sessions directory in
      Codex CLI, Export Chat in Cursor — which produces a markdown or text transcript. For an
      application form, upload that file as-is. To actually show it to a person, publish the
      .md to a page: it renders styled and readable at a private link, gated if needed, and
      the link updates in place when you export a better run.

```
# Session 2026-08-24T05:01:12

**Human:** Add standard RateLimit headers to the anonymous
publish endpoint, and Retry-After on the 429.

**Assistant:** Reading upload.ts to find the limiter... The
anon path counts rows in anon_uploads over a 24h window. I
will compute reset from the oldest row and emit the triplet
on every response, plus Retry-After when the cap is hit.

[tool_use: Read] {"file_path": "src/upload.ts", "offset":
60, "limit": 120}
[tool_result] 1059 lines truncated ...
[tool_use: Edit] {"file_path": "src/upload.ts", "old_str":
"if (anonUsed >= ANON_UPLOADS_PER_DAY) {", "new_str": ...
[tool_result] ok

**Assistant:** Tests: 14 passed. The 429 now carries
Retry-After computed from the window reset.
```

session.md · 214 KB · 4,816 linesAGENT SESSION · 24 AUGRate-limit headers for the anonymous APITHE ASKStandard RateLimit headers on anonymous publishes, Retry-After on the 429.Read upload.tsEdit ×114 tests passTHE FINISHReset computed from the oldest row in the window; the 429 now tells agents exactly when to come back.`curl -X POST api.stacktr.ee/sites -F file=@session.md`stacktr.ee/p/QYAfjS… · private · gate optionalPublished for real while writing this post — a raw .md in, a styled page out, one request.

## Why this suddenly matters

 Y Combinator's application added an optional question for the Spring 2026 batch that has been circulating again this weekend: "attach a coding agent session you're particularly proud of" — a .md or .txt file, up to 10MB. The form's own help text points at the mechanism: most coding agents have an `/export` command or an export button. The question is a tell about where evaluation is going: not "do you use AI tools" but "show us the actual session." And a culture of showing off sessions needs two things the tools half-provide: getting the transcript out, and making it readable.

## Getting it out, tool by tool

 As of this writing: Claude Code has `/export`, which writes the whole conversation — turns, tool calls, results — to a file or clipboard. Codex CLI keeps per-session transcripts under `~/.codex/sessions`. Cursor exports a chat from the conversation menu. Whatever the tool, the output is the same species: a long, honest, machine-formatted markdown file. For the YC form, that raw file is exactly what you upload; done.

## The wall problem

 Everywhere else, the export is where sharing goes to die. A real session runs thousands of lines: interleaved tool dumps, JSON arguments, retries. Sent as a file attachment, it gets skimmed for ten seconds. The fix is the same one that works for every other artifact an agent produces: put it on a page. We published this post's own test transcript while writing it — one request, no account:

```
`curl -X POST https://api.stacktr.ee/sites -F file=@session.md`
```

 The markdown renders as a styled page — headings, code blocks, readable width — at an unguessable link. Anonymous publishes live 24 hours; claim the page free to keep it, passcode it (any plan) if the session touches private code, and `update_site` replaces it in place when you export a better run, so the link you sent your cofounder on Monday shows Thursday's pick.

## Choosing the session worth showing

 Having reviewed a lot of our own: the impressive transcript is rarely the flawless one. It's the one with a real recovery — the agent hits a wall, diagnoses it, routes around it, and says what it verified. If you're picking between candidates, publish the shortlist to one gated link each and let your cofounder read them like documents rather than scroll them like logs. Prune repeated tool dumps if you're sharing publicly; keep the dead ends that show judgement. For applications, resist the rewrite — raw is the point, and reviewers can tell.

     See it in one call, no account

```
`curl -X POST https://api.stacktr.ee/sites -F file=@index.html
# → a private URL, live for 24h. Claim it free to keep it.`
```

If your agents emit HTML, they need a durable publish target: publish once, then
`update_site`
on every revision keeps the same URL forever. MCP, REST, and skills all speak it:
        the map is at [agent.txt](https://stacktr.ee/agent.txt).

[
Get an API key free
](https://app.stacktr.ee/?join=1&from=bridge_blog_export_coding_agent_session)
FAQ

## Frequent questions

     How do I export a session from Claude Code? +  Run /export inside the session. It writes the full conversation — every turn, tool call, and result — to a file (or the clipboard), as text or markdown. That file is what application forms and reviewers mean by "a session transcript."   How do I export a session from Codex or Cursor? +  Codex CLI keeps session transcripts on disk under ~/.codex/sessions, one file per session. Cursor exports a chat from the conversation menu (Export Chat). Most agent tools now have one of the two shapes: a slash command or an export button — YC’s own form notes exactly this. Details drift between versions, so check your tool’s current menu if neither appears where described.   What is the YC "coding agent session" application question? +  Starting with the Spring 2026 batch, Y Combinator’s application added an optional question: attach a coding agent session you are particularly proud of, as a .md or .txt file up to 10MB. It exists to let applicants show real fluency with AI coding tools rather than claim it. For the form itself you upload the raw file — no hosting involved.   How do I share an agent session with someone as a link? +  Publish the exported .md: POST it to api.stacktr.ee/sites (or use the publish tools any connected agent has) and the markdown renders as a styled, readable page at a private unguessable link. Add a passcode if it contains anything sensitive — that works on every plan — and the same URL updates in place if you re-export a better run.   Should I clean up a session transcript before sharing it? +  For an application, mostly no — the raw shape is the evidence, and reviewers reading transcripts can smell a rewrite. For sharing with a wider audience, prune the noise (repeated tool dumps, dead ends that add nothing) but keep the recovery moments: watching an agent hit a wall and route around it is usually the most impressive part of a real session.
Keep reading

## Related guides

   -  [ The design pass For HTML pages an agent shipped plain: one click to Stacktree’s design standard. ](/design-pass)
-  [ Why agents need a publish primitive The thesis: agent output deserves a URL, not an attachment. ](/blog/why-agents-need-a-publish-primitive)
-  [ For developers The publish API, MCP server, and the standing-report pattern. ](/developers)
-  [ What AI agents publish What actually gets published when agents hold the pen. ](/blog/what-ai-agents-publish)

References

## Sources and further reading

   -  [ Y Combinator application ↗ The application carrying the agent-session question (optional, .md/.txt, 10MB). ](https://www.ycombinator.com/apply)
-  [ Claude Code documentation ↗ Slash commands including /export. ](https://docs.anthropic.com/en/docs/claude-code)
-  [ OpenAI Codex CLI ↗ Session transcripts on disk under ~/.codex/sessions. ](https://github.com/openai/codex)

##  The session deserves better than an attachment.

 Publish the .md free, gate it if it's sensitive, update it in place when a better run lands.

 [ Sign up free → ](https://app.stacktr.ee/?join=1&from=cta_blog_export_coding_agent_session)       Private hosting for the HTML your agents make.

[](https://betalist.com/startups/stacktree?utm_campaign=badge-stacktree&utm_medium=badge&utm_source=badge-featured)[Featured on](https://devhunt.org/tool/stacktree)[](https://buildlist.io)[Get started free](https://app.stacktr.ee/?join=1)

## Product

- [Templates](/templates)
- [Example deliverables](/examples)
- [Custom domains](/custom-domains)
- [Client feedback](/client-feedback-loop)
- [One-time links](/one-time-view-links)
- [See how pages get read](/page-engagement)
- [Made with Stacktree](/made-with)
- [Security](/security)
- [Watch the demo](/demo)

## Agents

- [All integrations](/agents)
- [Claude Code](/claude-code)
- [OpenAI Codex](/codex)
- [Cursor](/cursor)
- [Claude.ai connector](/claude-ai-connector)
- [MCP server](/mcp-publish-html)
- [Deploy from Claude Code](/deploy-html-from-claude-code)
- [Skills](/skills)
- [Slack app](/slack)
- [n8n node](/n8n)
- [Agent payments (x402)](/x402)

## Alternatives

- [All comparisons](/alternatives)
- [Head-to-head comparisons](/compare)
- [Tiiny Host](/tiiny-host-alternative)
- [GitHub Pages (private)](/github-pages-private-alternative)
- [Vercel](/vercel-alternative-for-agents)
- [ngrok](/ngrok-alternative-for-html)
- [Display.dev](/display-dev-alternative)
- [Static.app](/static-app-alternative)
- [OpenAI Codex Sites](/openai-codex-sites-alternative)
- [here.now](/here-now-alternative)
- [Shippage](/shippage-ai-alternative)
- [Best private hosting](/best-private-html-hosting)

## Use cases

- [All use cases](/use-cases)
- [Share with clients](/share-with-clients)
- [Send a file to a client](/share-html-file-with-client)
- [Share Claude artifacts](/share-claude-artifacts)
- [Share Jupyter notebooks](/share-jupyter-notebook-html)
- [Host Storybook privately](/host-storybook-privately)
- [Architecture diagrams](/share-architecture-diagrams)
- [AI-generated reports](/host-ai-reports)
- [Internal HTML tools](/internal-tool-hosting)
- [Private HTML hosting](/private-html-hosting)
- [Vibe-coded page hosting](/vibe-coding-hosting)
- [Leave a website builder](/website-builder-migration)

## Learn

- [Blog](/blog)
- [Glossary](/glossary)
- [FAQ](/faq)
- [Agent-loop hosting](/agent-loop-hosting)
- [Why agents need a publish primitive](/blog/why-agents-need-a-publish-primitive)
- [MCP servers explained](/blog/mcp-servers-explained-for-developers)
- [What changed in the 2026-07 MCP spec](/blog/mcp-2026-spec-changes)
- [Sites in Codex explained](/blog/sites-in-codex-explained)
- [Private-by-default hosting](/blog/private-by-default-html-hosting)
- [An agent paid us $1 (x402)](/blog/agent-paid-to-provision-itself)
- [When a loop hits a paywall](/blog/loop-engineering-paywall)
- [Pricing](/pricing)
- [Self-host (new)](/self-host)
- [Changelog](/changelog)
- [Docs](https://stacktr.ee/docs)
- [About](/about)
© 2026 Stacktree · stacktr.ee

[Privacy](/privacy)[Terms](/terms)[Security](/security)[Dashboard](https://app.stacktr.ee)[npm](https://www.npmjs.com/package/stacktree-mcp)[Sitemap](/sitemap.xml)[llms.txt](/llms.txt)[API spec](/openapi.json)

---
Full markdown summary of the Stacktree marketing surface: https://stacktr.ee/llms-full.txt