#byhaushalt
LightNews LightNews
rotecodefraktion.bsky.social
Article 2 of the series: Plan Mode with #ClaudeCode on the #byhaushalt demo project. Shows concrete prompts for structural analysis of 10,000 pages of PDF, the review step before execution, and what distinguishes ADR from ordinary documentation. #ai #coding
www.rotecodefraktion.de/blog/agentic...
Erkunden statt loslegen — Plan Mode mit Subagents
Artikel 2 der Reihe — bevor byhaushalt seinen ersten Parser-Code sieht, kartiert Claude Code die PDF-Struktur der bayerischen Haushaltspläne. Plan Mode, Explore-Subagent und Architecture Decision Reco...
www.rotecodefraktion.de
May 10, 2026 at 9:22 PM
rotecodefraktion.bsky.social
Hooks: When tests run on autopilot — Article 9 of the series "#AgenticCoding with Claude Code".
Three hooks for #byhaushalt: format after every edit, block debug leftovers on commit, tests at turn end — but smart, not brute force.
#claudecode #ai
www.rotecodefraktion.de/en/blog/agen...
#byhauhalt
Hooks — Tests as a Gate in the Background
Article 9 of the series — three hooks for byhaushalt: PostToolUse formats after every edit, PreToolUse blocks commits with debug leftovers, Stop runs fast tests but only for changed layers. Why an ove...
www.rotecodefraktion.de
May 19, 2026 at 7:23 AM
rotecodefraktion.bsky.social
"Slash Commands — Automating Repetition Away" Article 4 of "Agentic Coding with Claude Code".
Three commands turn a manual parser run into a deterministic workflow over 16 Einzelpläne. xfail tests double as a candid status check.
#byhaushalt #llm #ai
www.rotecodefraktion.de/en/blog/agen...
Slash Commands — Automating Repetition Away
Article 4 of the series — three slash commands make the parser repeatable: /parse-epl for any Einzelplan, /check-totals with built-in pytest, /diff-vs-gesamt for comparison against the full budget. Pl...
www.rotecodefraktion.de
May 13, 2026 at 6:29 AM
rotecodefraktion.bsky.social
Article 8 of the series #AgenticCoding with #ClaudeCode: "End-to-End with #Playwright MCP, Tests That Claude Writes Itself"
Markdown spec in, Playwright test out. Claude opens the browser, walks the user flow, and writes the test.
#ai #byhaushalt #e2e
www.rotecodefraktion.de/en/blog/agen...
End-to-End with Playwright MCP — Tests That Claude Writes Itself
Article 8 of the series — Markdown spec as E2E test contract: Claude reads the user flow, explores live in the browser via Playwright MCP, finds robust selectors, and writes the Playwright test. Three...
www.rotecodefraktion.de
May 18, 2026 at 5:35 AM
rotecodefraktion.bsky.social
Article 6 of the series #Agentic Coding with #Claude Code: "Worktrees — Building Three Visualizations in Parallel"
Three Git worktrees, three dev servers, one dataset. Treemap, sunburst, and Sankey compared side by side in the browser.
#byhaushalt
www.rotecodefraktion.de/en/blog/agen...
Worktrees — Building Three Visualizations in Parallel
Article 6 of the series — three worktree branches, three visualizations over the same data model. Treemap, sunburst, and Sankey compared side by side rather than guessed at sequentially. Plus: why the...
www.rotecodefraktion.de
May 16, 2026 at 9:44 AM
rotecodefraktion.bsky.social
Article 7 of “Agentic Coding with #ClaudeCode”: Claude Code knows your libraries from training. But which version? configuring #MCP servers via .mcp.json so Context7 fetches current docs at request time — shown on a real shadcn/ui migration.
#byhaushalt #ai

www.rotecodefraktion.de/en/blog/agen...
MCP Servers — Documentation Lookup in the Editor
Article 7 of the series — two MCP servers, one project configuration: Context7 fetches current library documentation directly into the editor before Claude has to guess. Plus a shadcn/ui migration and...
www.rotecodefraktion.de
May 17, 2026 at 9:10 AM
rotecodefraktion.bsky.social
Three JSON files are not a data model. Article 5 of "Agentic Coding with Claude Code" shows how a plan file with three tasks turns into parallel subagents building schema, Parquet, and Hypothesis tests.
www.rotecodefraktion.de/en/blog/agen...

#ClaudeCode #AgenticCoding #byhaushalt
Subagent-Driven Development — Building a Data Model in Parallel
Article 5 of the series — v0.4 delivers three JSON files, not a data model. A plan file with three tasks and parallel subagents builds schema, normalization, and property-based tests. Plus: what forma...
www.rotecodefraktion.de
May 14, 2026 at 6:30 AM
rotecodefraktion.bsky.social
The code was done, the E2E tests green — and the numbers were still off.
The bonus article: validating the #byhaushalt parser against two independent sources. 5 bugfixes to hit the budget PDF to the cent. #agenticcoding won't replace developers.
#claudecode
www.rotecodefraktion.de/en/blog/agen...
Three Sources, One Truth — Validating the byhaushalt Parser Against PDF and STMFH Diagram
Bonus article of the series — how the byhaushalt demo parser holds up against two independent sources: the official EPL PDFs from the Bavarian finance ministry and the interactive budget diagram at st...
www.rotecodefraktion.de
May 23, 2026 at 9:37 AM
rotecodefraktion.bsky.social
MIT-licensed, with one exception: no use or distribution by the AfD or the right-wing populist media ecosystem (COMPACT, Junge Freiheit, PI-News, NIUS et al.). Open data, but not for those fighting the open society.
Repo: codeberg.org/rotecodefrak...

#noafd #fckafd #nonazis
byhaushalt
Ein Projekt, das den bayerischen Haushalt 2026/27 aus amtlichen PDFs auch maschinell und bearrierefrei lesbar machen soll und gleichzeitig als Demoprojekt für agentic coding dient.
codeberg.org
May 23, 2026 at 9:19 AM
rotecodefraktion.bsky.social
The byhaushalt demo is live: byhaushalt.rotecodefraktion.de
byhaushalt — Bayerischer Haushalt 2026/27
byhaushalt.rotecodefraktion.de
May 23, 2026 at 9:19 AM
rotecodefraktion.bsky.social
For nine articles we unpacked the #agenticcoding toolbox. The demo project still lacked one thing: a public URL.
Part 10 closes the gap — branch to live site. PR workflow, git worktree deploy, DNS CNAME, Let's Encrypt.
www.rotecodefraktion.de/en/blog/agen...
#claudecode #ai #byhaushalt #noafd
From Branch to Live URL — Worktree PR Workflow and Codeberg Pages Deploy
Article 10 of the series — how a Vite SPA in the repo becomes a live site on a custom subdomain. Branch protection for solo projects, a Bash script that fills the pages branch via git worktree, custom...
www.rotecodefraktion.de
May 23, 2026 at 9:18 AM
rotecodefraktion.bsky.social
Article 3: custom skills with pytest TDD on the #byhaushalt project. It shows why the trigger description is the most critical line in a skill — and how a sum consistency test prevents the parser from silently reading wrong values from the PDF.
#claudecode #ai
www.rotecodefraktion.de/en/blog/agen...
Writing Custom Skills — PDF Parser Skill with pytest TDD
Article 3 of the series — the first parser code for byhaushalt takes shape via TDD: a custom skill with trigger description, uv setup, six tests before implementation, and how the sum consistency test...
www.rotecodefraktion.de
May 12, 2026 at 6:09 AM
rotecodefraktion.bsky.social
Day 1 "Agentic Coding with #Claude Code" on www.rotecodefraktion.de/en/blog/agen... shows what that looks like in practice, on the byhaushalt demo project.
Sending an AI agent into a project means sending a colleague without memory. CLAUDE.md is the briefing file that changes that.
#ai #coding
CLAUDE.md as a Briefing
Article 1 of the series — we install Claude Code, set up the byhaushalt repo, write the first memory file, and pin down what the agent has to follow in every session. Plus: permissions, source data, t...
www.rotecodefraktion.de
May 10, 2026 at 10:34 AM