--- title: "Automating cross-repo documentation with GitHub Agentic Workflows" url: https://daily.dev/posts/automating-cross-repo-documentation-with-github-agentic-workflows-hftqrw7zb source_url: https://github.blog/ai-and-ml/github-copilot/automating-cross-repo-documentation-with-github-agentic-workflows type: article source: "GitHub Blog" published: 2026-07-08T21:13:21.414Z updated: 2026-07-09T04:21:47.741Z tags: ["github", "ai-agents", "github-actions"] reading_time: 11 upvotes: 18 comments: 2 language: 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. # Automating cross-repo documentation with GitHub Agentic Workflows **[GitHub Blog](https://daily.dev/sources/ghblog)** · 11 min read · 18 upvotes · 2 comments ## Summary The Aspire team (microsoft/aspire) built a cross-repo documentation automation pipeline using GitHub Agentic Workflows. When a product pull request merges, a workflow called pr-docs-check automatically drafts a documentation pull request in the separate aspire.dev repo, assigns the original feature's SME as reviewer, and posts a link back to the source PR. The system uses a scoped GitHub App token limited to exactly two repos, a safe-outputs handler that separates agent reasoning from write actions, and milestone-to-release-branch mapping for accurate targeting. Over a 30-day window spanning Aspire 13.3 and 13.4, 82 docs PRs were created with a 100% merge rate and median time-to-merge of 44.8 hours. The post covers the security model, lessons learned (over-eager docs detection, cross-repo checkout patterns, prompt budget management), and four additional companion workflows also running in production. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Community discussion Top comments from developers on daily.dev. **@kartiknvj** · 1 upvotes > The design detail worth stealing here is the safe-outputs handler that separates the agent's reasoning from the write actions, so the model proposes and a constrained handler executes. Paired with a token scoped to exactly two repos, that is a much saner security posture than giving an agent broad write access and hoping the prompt holds. The 100 percent merge rate over 82 PRs is less about the model being clever and more about keeping the task narrow and the human SME in the loop as reviewer, which is the pattern I trust for putting agents in production. ## Similar posts on daily.dev - [Streamline Aspire SDK Updates with GitHub Actions](https://daily.dev/posts/streamline-aspire-sdk-updates-with-github-actions-3t4ynnclc) · We Are .NET · 0 upvotes · 0 comments - [Automate Astro Upgrades with GitHub Agentic Workflows](https://daily.dev/posts/automate-astro-upgrades-with-github-agentic-workflows-aow2uvryc) · We Are .NET · 0 upvotes · 0 comments - [How We Used Agentic Pipelines to Swarm Documentation](https://daily.dev/posts/how-we-used-agentic-pipelines-to-swarm-documentation-sdly2qldb) · Atlassian · 0 upvotes · 0 comments --- Tags: [#github](https://daily.dev/tags/github), [#ai-agents](https://daily.dev/tags/ai-agents), [#github-actions](https://daily.dev/tags/github-actions) [View this post on daily.dev](https://daily.dev/posts/automating-cross-repo-documentation-with-github-agentic-workflows-hftqrw7zb)