Run PDF and API documentation builds in parallel

[konard]

Summary

Resolve #59 by moving the expensive PDF build off the API-documentation critical path.

• run PDF generation and DocFX API generation concurrently on separate GitHub-hosted runners
• upload the PDF and HTML site as independent artifacts
• make the publishing job join, download, and validate both artifacts before deploying
• validate the fully assembled site on pull requests without publishing it
• require that pull-request assembly in the terminal pipeline gate so a skipped validation cannot pass silently
• keep .github/scripts/publish-csharp-docs.sh as the single repository-owned Pages publisher, preserving other content on gh-pages
• merge the latest main and remove the draft's unavailable pdf-generation self-hosted label and duplicate publisher

Publication now waits for max(PDF build, API build) instead of PDF build + API build, while still refusing to deploy an incomplete site.

Reproduction and regression coverage

Before the fix, buildDocumentation declared generatePdfWithCode as a dependency, so DocFX could not start until the slow TeX job finished. The updated workflow-policy test reproduced this dependency and failed before the workflow change.

The regression tests now verify that:

• both build jobs depend only on documentation-input detection
• both jobs produce their own named artifact
• the publisher depends on both jobs and downloads both artifacts
• pull requests download and validate the assembled site without deploying it
• the terminal gate rejects a skipped pull-request assembly when documentation changed

Verification

• node --test .github/scripts/*.test.mjs — 31 passed
• .github/scripts/validate-csharp-package.sh with .NET 8 — clean build, 2 tests passed, package and embedded symbols validated
• actionlint 1.7.12 — passed
• shellcheck .github/scripts/*.sh — passed
• DocFX 2.80.1 with --warningsAsErrors — 0 warnings, 0 errors
• .github/scripts/generate-csharp-pdf.sh — generated a valid eight-page PDF
• git diff --check origin/main...HEAD — passed
• C# CI run 35691240686 — every required job passed, including both documentation builders, assembly validation, and the terminal gate
• Codacy Static Code Analysis on f3ac75a — passed

Resolves #59

[konard]

Double check everything delivered with highest possible quality, resolve conflicts.

[konard]

🤖 AI Work Session Started

Starting automated work session at 2026-09-22T04:52:52.836Z

The PR has been converted to draft mode while work is in progress.

This comment marks the beginning of an AI work session. Please wait for the session to finish, and provide your feedback.

Runtime: solve v2.29.0 · tool codex · model gpt-5.6-sol · task image konard/hive-mind-dind:2.29.0@sha256:11a236d03854cbaea86fdd7ca06265c912cf0b67d1fcc9b3eb60bd84c43ff749

[konard]

Codacy needs a maintainer/service retry on the final commit 1f6230d: its source-preparation stage succeeded, but Codacy’s internal Diff and Deltas stages errored and returned ACTION_REQUIRED with zero annotations. The public Codacy report says there are no new or fixed issues, and the immediately preceding analysis (718708d) passed with “Codacy found no issues in your code.” GitHub denied this account permission to re-request the app-owned check (HTTP 404). All repository-owned workflows and CodeQL pass on 1f6230d.

[konard]

🤖 Solution Draft Log

This log file contains the complete execution trace of the AI solution draft process.

💰 Cost estimation:

• Model: GPT-5.6 Sol
• Provider: OpenAI
• Public pricing estimate: $8.391847

📊 Context and tokens usage:

• 248.7K / 200K (124%) input tokens, 27.8K / 128K (22%) output tokens

Total: (248.7K + 7.0M cached) input tokens, 27.8K output tokens, $8.391847 cost

🤖 Models used:

• Tool: OpenAI Codex
• Requested: gpt-5.6-sol
• Thinking level: xhigh (~31999 tokens)
• Model: GPT-5.6 Sol (gpt-5.6-sol)

📎 Log file uploaded as Gist (3161KB)

• View complete solution draft log

---

Now working session is ended, feel free to review and add any feedback on the solution draft.

[konard]

🔄 Auto-restart 1/5

Reason: CI failures detected

Starting new session to address the issues.

---

Auto-restart-until-mergeable mode is active. This run will stop after 5 restart iterations in total.

[konard]

Working session summary

Fixed and finalized PR #131.

• Codacy now passes on commit f3ac75a.
• Added regression coverage requiring documentation assembly validation on pull requests.
• All applicable GitHub Actions and CodeQL checks pass.
• Local Node, .NET 8, ShellCheck, Actionlint, and whitespace checks pass.
• Branch includes latest main, is clean and mergeable.
• PR title/description are updated and the PR is ready for review.

---

This summary was automatically extracted from the AI working session output.

[konard]

🔄 Auto-restart-until-mergeable Log 1/5

This log file contains the complete execution trace of the AI solution draft process.

💰 Cost estimation:

• Model: GPT-5.6 Sol
• Provider: OpenAI
• Public pricing estimate: $2.244240

📊 Context and tokens usage:

• 82.0K / 200K (41%) input tokens, 12.8K / 128K (10%) output tokens

Total: (82.0K + 1.5M cached) input tokens, 12.8K output tokens, $2.244240 cost

🤖 Models used:

• Tool: OpenAI Codex
• Requested: gpt-5.6-sol
• Model: GPT-5.6 Sol (gpt-5.6-sol)

📎 Log file uploaded as Gist (4417KB)

• View complete solution draft log

---

Now working session is ended, feel free to review and add any feedback on the solution draft.

[konard]

🎉 Auto-merged

This pull request has been automatically merged by hive-mind.

• All CI checks have passed

---

Auto-merged by hive-mind with --auto-merge flag