Mode B deployer nu algo-settings automatisk til alle nye projekter.
Frederick Winslow Taylor — videnskabelig arbejdsledelse, 'one best way'.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- BCQuality: Kaoru Ishikawa — kvalitetscirkler, fiskebensdiagram,
'kvalitet er alles ansvar'
- BC-MCP: Grace Hopper — foerste compiler, COBOL, debugging-begrebet
- AL-Complexity: Eliyahu M. Goldratt — Theory of Constraints,
The Goal, kritisk kaede
- The Court: Platons Akademi — grundlagt 387 f.Kr., 900 aars
virke, metoden frem for svaret
- M365: Alexander Graham Bell — telefonen, AT&T, forbindelsen
paa tvaers af afstand
AL-Triage (Larrey) og ALGo-Settings (Taylor) afventer bekraeftelse.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Hver agent faar nu et 'Who I Am'-afsnit med fuldt navn,
levedatoer og de vigtigste bedrifter fra det virkelige liv:
- Columbo: LAPD-detektiv, hver sag loesat, ingen efternavn
- Florence Nightingale: sygepleje, 42%->2% doedelighed, statistik
- Francis Bacon: empirisk metode, Novum Organum 1620
- Immanuel Kant: Kategorisk Imperativ, Critique of Pure Reason
- Abraham Lincoln: 16. prasident, Borgerkrig, afskaffelse af slaveriet
- Marcus Aurelius: Roms 16. kejser, Meditations, stoisk reduktion
- Charlie Munger: Berkshire Hathaway, mental models, inversion
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Den hyppigste setup-fejl i pre-CURABIS-Standard projekter:
.mcp.json peger paa en projektlokal bc-mcp-bridge.js der ikke
eksisterer i stedet for ~/.claude/bc-mcp-bridge.js.
Mode B tjekker nu businesscentral-stien og retter den stille
hvis den er forkert. Rapporterer aendringen til udvikleren.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
florence.agent.md:
- Tilfoej Step 0: timestamp-gate
Laes ~/.claude/.florence-timestamp — spring runden over
hvis under 30 minutter gammel. Skriv timestamp efter runde.
Forhindrer Florence i at kore ved hver session-aabning.
curabis-standard.agent.md:
- CLAUDE.md-template: erstat 24 URL-fetches med local cache-laesning
fra ~/.claude/bcquality-knowledge/ (architecture/ testing/ mcp/)
Session-start reduceres fra ~25 netvaerkskald til 0.
Global ~/.claude/CLAUDE.md opdateret separat:
- Auto-update downloader nu knowledge-filer til cache ved SHA-aendring
- Session-start laeder fra cache, ikke fra URLs
- Florence korer kun hvis >= 30 min siden sidst
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Fix 1: Mode B laes agent frontmatter description og foreslaer
praecis CLAUDE.md-tekst for manglende agenter
Fix 3: Mode B substituerer {PROJECT_NAME} og {SETUP_DATE} i
HEARTBEAT.md fra CLAUDE.md-overskrift eller git remote
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Observation fra session 2026-06-25: agenter installeret via Mode B forbliver
usynlige for Claude fordi CLAUDE.md ikke opdateres automatisk.
Foreslaaet af Francis. Valideret af Immanuel (alle 4 tests bestaaet).
Afventer Michaels godkendelse via merge.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Three-judge appellate court for BCQuality governance. Convenes Lincoln
(essential question), Aurelius (Stoic reduction), and Munger (inversion)
to deliberate on the strategic health of the BCQuality rulebook.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- All 7 existing Tier 1/2 knowledge files now include a BCApps Reference
section with concrete source links and observed patterns
- New: bcpt-scenarios-must-be-app-specific — PerformanceTest apps must
include app-domain BCPT scenarios, not only Microsoft generic samples
- New: permission-sets-must-follow-least-privilege — View/Edit/Admin
hierarchy with IncludedPermissionSets, mirroring BCApps BusFound pattern
- api-page-key-fields-must-be-editable-on-insert clarified: SystemId as
ODataKeyField + Editable=false is valid (auto-generated); rule applies
to consumer-provided key fields only
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Users refer to tasks by taskNo in conversation (e.g. 'opgave 51').
Commit messages must use taskId (e.g. 8738) — different field.
Added explicit Gotcha block and sharpened step 3 in How to find.
Approved by: mid (Michael Dieringer)
Francis observes BC-MCP session patterns and identifies where existing rules
are too superficial (Type A sharpening) or where no rule covers the pattern
(Type B gap). Type B gaps are handed to Immanuel for Categorical Imperative
validation before entering BCQuality.
Immanuel guards the knowledge base by running four tests (universalizability,
project-specificity, clarity, additive value) before any rule is approved.
When a developer asks for open tasks, only return tasks from the project(s)
linked to the current git repository. Flag it if no project is linked.
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
