admin/ki-buch
1
0 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
main
ki-buch/anhang/E-stichwortverzeichnis.md
T
grok-bot eba5a63528
CI / Companion (py3.11) (push) Has been cancelled
Details
CI / Companion (py3.12) (push) Has been cancelled
Details
Hands-on Companion first-class: Kap. 10, Fake-MCP und examples-Layout. 
Kapitel 10 mit 5-Min-Start, Lab L0–L5 und Keep/Throw schärfen; Companion um
patterns/, golden Traces, Policy-CI und examples/-Migration erweitern. PDF und
Lesepfade/Exec-Summaries an den neuen Laborstandard anbinden.
2026-07-26 12:46:45 +02:00

7.4 KiB
Raw Permalink Blame History

Anhang E: Stichwortverzeichnis

Verweise auf Kapitel (Hauptstelle fett). Abschnittnummern entsprechen dem Manuskriptstand nach Harness- und Projekt-Information-Erweiterung.

A

  • Agent – 4, 5, 7, 8
  • Agent Infra Owner – 13.4
  • AGENTS.md / CLAUDE.md – 5.12, 10.13
  • Alignment – 1.2, 3.5
  • Allowlist / Denylist – 7.8, 11.7, 10.7.5
  • Anti-Hype-Checkliste – 13.6, Schlusswort
  • Architektur- und Musterbuch (Positionierung) – Vorwort, Einleitung, Anhang D
  • Autonomiegrad (L0–L3) – 7.4, 11.9, 12.4, 13.5
  • Autoregression – 2.1

B

  • Best Practices (Kapitelenden) – 1–13
  • Blast Radius – 11, 9.7
  • Budget (Tokens/Steps) – 3.7, 5.5, 8.5

C

  • CA-MCP – 5.8, 8.6
  • Capability Negotiation – 4.7.1
  • CI/CD-Integration – 7.6, 9, 10.7.4
  • Companion (code/, Tags book-2026-07/latest) – Vorwort, 10, code/COMPATIBILITY.md, Anhang D
  • Context Engineering – Einleitung, 3.7, 7.8, 10.8, 12.8
  • Context Engineering Owner – 13.4
  • Context Pack / Working Set – 7.8
  • Context Quality Management – 11.8
  • Context Trace / TRACE-DEMO-1 – 7.8, 10.8.6, 11.7.7, Anhang C.12
  • Executive Summaries – teil-*/00-executive-summary.md
  • MCP Spec 2025-11-25 – Vorwort, 2.6, 4.7, Anhang D.1
  • Context Window – 2.3, 3.6
  • Context-Lifecycle / Compact / Session – 5.12.5, 10.13.6
  • Recovery Pack – 5.12.5, 10.13.6
  • Coding Agent – 4.5, 4.6, 7, 10
  • codebase-memory-mcp / Code Knowledge Graph – 5.9, 5.12.3, 10.7.2, 10.13
  • CrewAI – 6.4, 6.8
  • Cursor (IDE-Klasse) – 4.5, 9.6

D

  • Data Layer (MCP) – 2.6, 4.7
  • Definition of Done / Ready – 7, 8.3
  • Demo-Workspace – 10, code/05-coding-agent/demo_workspace
  • Discovery – 4.7, 5.8
  • Don’t blame the model – 5.11, 11.10.6
  • DORA / State of AI-assisted Software Development – 13.1, Anhang F.2, Anhang D.6a

E

  • Eval / Evaluation – 6.5, Anhang C, Anhang F
  • Embedding – 2.1
  • Empirische Evidenz / Feldstudien – Anhang F, 13.1, 6.5.6
  • Evidenzgrad (Studie / Faustregel / Incident) – Anhang F

F

  • FastMCP – 2.6, 10.7.1, code/01–04
  • Anhang F (Empirische Evidenz 2026) – F, Verweise 3, 6.5, 8.6, 11, 13
  • Function Calling – 1.3, 4.7.3
  • Feldstudie / RCT (Copilot, METR) – 13.1, Anhang F.2
  • Frühindikatoren – 12.2, 12.4, 12.7
  • Führungsentscheidungen (90 Tage) – 13.5, 7.9

G

  • Gateway – 6.8, 9, 11.7
  • Git-Historie / git-mcp – 7.8.4, 10.7.4
  • Graph-Orchestration / LangGraph – 5.5, 6.3
  • Grounding – 3.2, 6.5

H

  • Halluzination – 3.2
  • Handoff – 5.10, 8.6
  • Harness / Harness Engineering – 4.5, 5.11, 11.10
  • Hybride Projekt-Information – 5.12.4, 10.13
  • HITL – 7, 8, 11
  • Hooks – 5.11, 10.12.6, 11.10.3
  • Host / Client / Server – 2.6.2, 4.7.1

I

  • IDE-Integration – 9.6
  • Impact Analysis – 5.9, 5.12.3, 10.7.2
  • Indirect Prompt Injection – 11.7.4, Anhang C
  • Invest-Priorisierung (12 Monate) – 12.5, Schlusswort
  • Issues / Tickets – 7.8.5, 10.7.4

J

  • JSON-RPC – 4.7.1
  • Just-in-Time Tools – 3.7, 5.3

K

  • Knowledge Graph / codebase-memory – 5.9, 5.12, 10.7.2, 10.13
  • KV-Cache – 2.2

L

  • LangGraph – 6.3, 6.8
  • Least Privilege – 11.7
  • LLM-Evolution – 1
  • Local vs Remote MCP – 9.7, 11.7.2
  • Lost in the Middle – 2.2, 3.2

M

  • MCP – Einleitung, 2.6, 4.7, 5–13, Anhänge
  • MCP-Bench / MCP-AgentBench / MCPSecBench – 6.5.6, Anhang F.3
  • MCP Apps / Extensions / Roadmap – 12.6
  • MCP Go/No-Go – 12.6, 4.7.5
  • MCP Server Owner – 13.4
  • Memory (Agent) – 3.4, 5.9, 10.7.3
  • METR (OSS-Dev-Produktivität 2025) – 13.1, Anhang F.2
  • Multi-Agent – 5.10, 8, 8.6, Anhang F.4

O

  • Observation Compression – 10.8.4
  • On-Demand Loading – 3.7, 10.8.1
  • Open-Source vs Closed Models – 1.4
  • Orchestrator / Supervisor – 5.2, 8

P

  • Plan-and-Execute – 4.3
  • Policy as Code – 11.8, 11.10.2
  • Production Harness – 11.10
  • Progressive Disclosure – 5.4, 5.12, 10.12, 10.13
  • Project Memory – 5.9, 10.7.3, 10.13
  • Prompts (MCP) – 4.7.2, 10.8.3, 10.12.5
  • Projekt-Information – 5.12, 10.13
  • Provenance – 3.6, 3.7

R

  • RAG – 2.6.1, 5.12.3
  • ReAct – 4.3
  • Redaction – 3.7, 11
  • Resources (MCP) – 4.7.2
  • Review-Agenten – 7.5
  • Review-Engpass – 13.2
  • Risiko multiplikativ – 11.1
  • Rollen (Agenten-Org) – 8.2, 13.4
  • Rollen (Hüte vs. Stellen) – 13.4

S

  • Sampling / Temperature – 2.4
  • Sandboxing – 10.7.5, 11.7.5
  • Scorecard (90 Tage) – 7.9, 13.5, Anhang C.2
  • SDLC – 7
  • Schlusswort (Leitmotive × Entscheidung) – Schlusswort
  • Selective Resource Exposure – 10.8.2
  • Shared Context Store – 5.8, 8.6
  • Skills (Agent Skills) – 5.4, 4.5, 10.12
  • Skill Engineering – 10.12
  • Security Checklist MCP – 11.7.8
  • STDIO / Streamable HTTP – 2.6, 4.7, 9.7
  • Stabilitäts-Matrix (stabil / mittelstabil / volatil) – Vorwort, Einleitung, Kap. 2/4–11, Anhang D
  • Structured Output – 2.4
  • Supply Chain – 11.7.6
  • System Prompt – 2.5, 4.5.5, 10.12.2 · Abgrenzung Skill/MCP-Prompt/AGENTS.md – Anhang A

T

  • Team-Readiness-Checkliste – 11 (Ende)
  • Threat-Model-Template – 11.7.0
  • Tools (MCP) – 4.7.2, 10.7 · Spec 2025-11-25
  • Token / Tokenizer – 2.1
  • Tool-Bloat – 2.6, 4.7
  • TRACE-DEMO-1 / Context Trace – 10.8.6, 3, 11.7.7, Anhang C.12
  • Trust Boundary – 11.7.1

U

  • Untrusted Data – 11.7.4
  • USB-C-Analogie – Vorwort, 2.6.2

V

  • Verifier – 8, 5
  • Vendor-Lock-in – 1.4, 6, 13

W

  • Wann MCP nicht – 4.7.5, 12.6
  • Wetten-Matrix – 12.4, Schlusswort
  • Working Set – 7.8.2

Z

  • Zukunftsszenarien A–E + Frühindikatoren – 12.7
  • 30-Minuten-Lesepfad – README · Exec Summaries teil-*/00-executive-summary.md

Datei- und Codeindex

Artefakt Ort
Companion-Root code/ (README, COMPATIBILITY, CHANGELOG, pyproject)
CI Smoke-Tests code/tests/
Minimal-Server code/01-minimal-server/
Project-Memory-Server code/02-project-memory/
Codebase-Graph-Server code/03-codebase-graph/
Dev-Tools-Server code/04-dev-tools/
Coding-Agent-Demo code/05-coding-agent/
Core-System-Prompt code/05-coding-agent/prompts/core_system.md
Skills code/05-coding-agent/skills/*/skill.md
Root-AGENTS.md (Demo) code/05-coding-agent/demo_workspace/AGENTS.md
Domain-Override code/05-coding-agent/demo_workspace/src/billing/AGENTS.md
Agent-Index / Recovery Pack code/05-coding-agent/demo_workspace/docs/agent/
ADR-Beispiel code/05-coding-agent/demo_workspace/docs/adr/
Host-Config-Beispiel code/host-config.example.json
Diagramm-Inventar bilder/README.md

Stichwortverzeichnis moniert die Lehrstruktur; für Print-Produktion Seitenzahlen nach Layout ergänzen.

Reference in new issue View Git Blame Copy Permalink