santismm-knowledge-mcp 0.2.1 → 0.3.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (39) hide show
  1. package/README.md +35 -15
  2. package/content/architectures/ai-workforce.json +3 -1
  3. package/content/claims/a-harness-can-degrade-the-agent.json +51 -0
  4. package/content/claims/harness-changes-outcomes.json +57 -0
  5. package/content/claims/harness-engineering-is-a-distinct-discipline.json +43 -0
  6. package/content/claims/harness-is-the-durable-asset.json +43 -0
  7. package/content/claims/maturity-levels-are-ordered-by-dependency.json +43 -0
  8. package/content/claims/practice-converges-on-eight-components.json +58 -0
  9. package/content/claims/verified-outcome-is-the-unit.json +50 -0
  10. package/content/governance/agentic-ai-governance-checklist.json +3 -1
  11. package/content/harness/HRN-001-definition-and-overview.es.md +6 -3
  12. package/content/harness/HRN-001-definition-and-overview.md +6 -3
  13. package/content/harness/HRN-001-definition-and-overview.pt.md +6 -3
  14. package/content/harness/HRN-013-glossary.es.md +1 -1
  15. package/content/harness/HRN-013-glossary.md +1 -1
  16. package/content/harness/HRN-013-glossary.pt.md +1 -1
  17. package/content/knowledge/foundation-models.json +2 -1
  18. package/content/knowledge/harness-engineering.json +3 -3
  19. package/content/library/the-agentic-enterprise-needs-an-immune-system.md +16 -0
  20. package/content/library/the-stopwatch-and-the-exam.md +5 -1
  21. package/content/matrix/agentic-control-matrix.json +408 -5
  22. package/content/maturity/harness-maturity-model.json +569 -0
  23. package/content/patterns/correlated-run-trace.json +329 -0
  24. package/content/patterns/egress-allowlist.json +4 -3
  25. package/content/patterns/evaluator-optimizer.json +1 -0
  26. package/content/patterns/goal-decomposition.json +1 -0
  27. package/content/patterns/human-approval-gate.json +1 -0
  28. package/content/patterns/long-term-memory.json +1 -0
  29. package/content/patterns/orchestrator-workers.json +1 -0
  30. package/content/patterns/output-boundary-encoding.json +320 -0
  31. package/content/patterns/recovery-strategy.json +8 -3
  32. package/content/patterns/sandboxed-execution.json +1 -0
  33. package/content/patterns/supervisor-agent.json +1 -0
  34. package/content/scorecard/harness-scorecard.json +460 -0
  35. package/dist/articles.js +120 -0
  36. package/dist/content.js +14 -1
  37. package/dist/shape.js +150 -3
  38. package/dist/tools.js +202 -4
  39. package/package.json +1 -1
@@ -35,7 +35,7 @@ The following are the canonical definitions of terms used throughout this corpus
35
35
  - **Agent:** a system that uses a language model in a loop to pursue a goal, deciding which actions (tool calls) to take based on observations until a stopping condition is met.
36
36
  - **Agentic system:** a software system whose behavior is driven by one or more agents, including all the surrounding scaffolding required to make them reliable.
37
37
  - **Harness:** the engineered scaffolding around a model — memory, tools, planning, orchestration, observability, evaluation, governance, and security — that turns a raw model into a reliable agentic system.
38
- - **Harness Engineering:** the discipline responsible for building reliable agentic systems for enterprise environments by designing and operating the harness.
38
+ - **Harness Engineering:** the emerging discipline responsible for building reliable agentic systems for enterprise environments by designing and operating the harness.
39
39
  - **Model / LLM:** the underlying large language model that performs reasoning and generation; in the harness it is treated as a powerful but non-deterministic, manipulable component.
40
40
  - **Autonomy gradient:** the spectrum of control modes from fully autonomous to human-in-the-loop, assigned per action class by policy.
41
41
  - **Blast radius:** the maximum harm an action (or a compromised agent) can cause; a core quantity to bound.
@@ -20,7 +20,7 @@ Os nomes ingleses de uso estabelecido na indústria (*harness*, *tool*, *span*,
20
20
  - **Agente:** um sistema que usa um modelo de linguagem dentro de um ciclo para perseguir um objetivo, decidindo que ações (chamadas a ferramentas) tomar a partir das observações até se cumprir uma condição de paragem.
21
21
  - **Sistema agêntico:** um sistema de software cujo comportamento é conduzido por um ou vários agentes, incluindo todo o andaime circundante necessário para os tornar confiáveis.
22
22
  - **Harness:** o andaime de engenharia em torno de um modelo — memória, ferramentas, planejamento, orquestração, observabilidade, avaliação, governança e segurança — que converte um modelo em bruto num sistema agêntico confiável.
23
- - **Engenharia de Harness:** a disciplina responsável por construir sistemas agênticos confiáveis para ambientes empresariais, desenhando e operando o harness.
23
+ - **Engenharia de Harness:** a disciplina emergente responsável por construir sistemas agênticos confiáveis para ambientes empresariais, desenhando e operando o harness.
24
24
  - **Modelo / LLM:** o grande modelo de linguagem subjacente que raciocina e gera; dentro do harness é tratado como um componente poderoso mas não determinista e manipulável.
25
25
  - **Gradiente de autonomia:** o espectro de modos de controle, de totalmente autônomo a humano no ciclo, atribuído por política a cada classe de ação.
26
26
  - **Raio de impacto (*blast radius*):** o dano máximo que uma ação, ou um agente comprometido, pode causar; uma grandeza central a limitar.
@@ -4,7 +4,8 @@
4
4
  "updated": "2026-06-21",
5
5
  "version": "1.0",
6
6
  "featured": true,
7
- "related": ["fine-tuning", "embeddings", "agentic-ai", "harness-engineering"],
7
+ "related": ["fine-tuning", "embeddings", "agentic-ai", "harness-engineering",
8
+ "reasoning-models"],
8
9
  "references": [
9
10
  { "title": "Bommasani et al. — On the Opportunities and Risks of Foundation Models (2021)", "url": "https://arxiv.org/abs/2108.07258" },
10
11
  { "title": "Vaswani et al. — Attention Is All You Need (2017)", "url": "https://arxiv.org/abs/1706.03762" }
@@ -18,7 +18,7 @@
18
18
  "locales": {
19
19
  "en": {
20
20
  "title": "What is Harness Engineering?",
21
- "summary": "Harness engineering is the discipline of designing and optimizing the scaffolding around an AI model — the prompts, tools, memory, environment, control loop and guardrails — so the model performs reliably on real tasks. Its core premise: as base models converge in raw capability, competitive advantage shifts from the model itself to the harness built around it. The same model can pass or fail a task depending almost entirely on its harness.",
21
+ "summary": "Harness engineering is the emerging discipline of designing and optimizing the scaffolding around an AI model — the prompts, tools, memory, environment, control loop and guardrails — so the model performs reliably on real tasks. Its core premise: as base models converge in raw capability, competitive advantage shifts from the model itself to the harness built around it. The same model can pass or fail a task depending almost entirely on its harness.",
22
22
  "definition": "Harness engineering is the practice of designing, building and optimizing the scaffolding (tools, memory, prompts, environment and control loop) that turns a model's raw capability into reliable, goal-directed action.",
23
23
  "takeaways": [
24
24
  "The harness is everything around the model that converts capability into action.",
@@ -64,7 +64,7 @@
64
64
  },
65
65
  "es": {
66
66
  "title": "¿Qué es la Ingeniería de Harness (Harness Engineering)?",
67
- "summary": "La ingeniería de harness es la disciplina de diseñar y optimizar el andamiaje alrededor de un modelo de IA —prompts, herramientas, memoria, entorno, bucle de control y guardarraíles— para que el modelo rinda de forma fiable en tareas reales. Su premisa central: a medida que los modelos base convergen en capacidad bruta, la ventaja competitiva se desplaza del modelo al harness que lo rodea. El mismo modelo puede aprobar o fallar una tarea casi por completo según su harness.",
67
+ "summary": "La ingeniería de harness es la disciplina emergente de diseñar y optimizar el andamiaje alrededor de un modelo de IA —prompts, herramientas, memoria, entorno, bucle de control y guardarraíles— para que el modelo rinda de forma fiable en tareas reales. Su premisa central: a medida que los modelos base convergen en capacidad bruta, la ventaja competitiva se desplaza del modelo al harness que lo rodea. El mismo modelo puede aprobar o fallar una tarea casi por completo según su harness.",
68
68
  "definition": "La ingeniería de harness es la práctica de diseñar, construir y optimizar el andamiaje (herramientas, memoria, prompts, entorno y bucle de control) que convierte la capacidad bruta de un modelo en acción fiable y dirigida a objetivos.",
69
69
  "takeaways": [
70
70
  "El harness es todo lo que rodea al modelo y convierte capacidad en acción.",
@@ -110,7 +110,7 @@
110
110
  },
111
111
  "pt": {
112
112
  "title": "O que é Engenharia de Harness (Harness Engineering)?",
113
- "summary": "A engenharia de harness é a disciplina de projetar e otimizar o andaime ao redor de um modelo de IA — prompts, ferramentas, memória, ambiente, laço de controle e guard-rails — para que o modelo tenha desempenho confiável em tarefas reais. Sua premissa central: à medida que os modelos base convergem em capacidade bruta, a vantagem competitiva se desloca do modelo para o harness à sua volta. O mesmo modelo pode passar ou falhar numa tarefa quase inteiramente conforme seu harness.",
113
+ "summary": "A engenharia de harness é a disciplina emergente de projetar e otimizar o andaime ao redor de um modelo de IA — prompts, ferramentas, memória, ambiente, laço de controle e guard-rails — para que o modelo tenha desempenho confiável em tarefas reais. Sua premissa central: à medida que os modelos base convergem em capacidade bruta, a vantagem competitiva se desloca do modelo para o harness à sua volta. O mesmo modelo pode passar ou falhar numa tarefa quase inteiramente conforme seu harness.",
114
114
  "definition": "A engenharia de harness é a prática de projetar, construir e otimizar o andaime (ferramentas, memória, prompts, ambiente e laço de controle) que converte a capacidade bruta de um modelo em ação confiável e orientada a objetivos.",
115
115
  "takeaways": [
116
116
  "O harness é tudo o que rodeia o modelo e converte capacidade em ação.",
@@ -0,0 +1,16 @@
1
+ ---
2
+ title: "The agentic enterprise needs an immune system"
3
+ titles: { es: "La empresa agéntica necesita un sistema inmunitario", pt: "A empresa agêntica precisa de um sistema imunológico" }
4
+ date: "2026-08-28"
5
+ kind: "article"
6
+ summary: "How to design AI agents that can act without concentrating trust: identity, least privilege, containment, oversight, and recovery."
7
+ summaries: { es: "Cómo diseñar agentes de IA que puedan actuar sin concentrar toda la confianza: identidad, mínimo privilegio, contención, supervisión y recuperación.", pt: "Como projetar agentes de IA capazes de agir sem concentrar toda a confiança: identidade, privilégio mínimo, contenção, supervisão e recuperação." }
8
+ topics: ["Agentic AI", "AI Security", "AI Governance"]
9
+ featured: true
10
+ host: "articles.santismm.com"
11
+ canonicalUrl: "https://articles.santismm.com/the-agentic-enterprise-needs-an-immune-system/"
12
+ links:
13
+ - { platform: "site", lang: "en", url: "https://articles.santismm.com/the-agentic-enterprise-needs-an-immune-system/" }
14
+ - { platform: "site", lang: "es", url: "https://articles.santismm.com/la-empresa-agentica-necesita-un-sistema-inmunitario/" }
15
+ - { platform: "site", lang: "pt", url: "https://articles.santismm.com/a-empresa-agentica-precisa-de-um-sistema-imunologico/" }
16
+ ---
@@ -1,12 +1,16 @@
1
1
  ---
2
2
  title: "The Stopwatch and the Exam"
3
+ titles: { es: "El cronómetro y el examen", pt: "O cronômetro e a prova" }
3
4
  date: "2026-06-15"
4
5
  kind: "article"
5
- summary: "Static benchmarks did not die. They stopped being enough. The move from measuring capability to measuring agency, the harness problem underneath it, and where attention belongs now. Backed by a public catalog of 69 agentic benchmarks."
6
+ summary: "Static AI benchmarks did not die; they stopped being enough. A grounded look at the move from capability to agency, the harness problem, and where to put attention now. Backed by a public catalog of 69 agentic benchmarks."
7
+ summaries: { es: "Los benchmarks estáticos no han muerto, pero ya no bastan. Del examen a la agencia: arneses, duración, coste, repetición y valor económico.", pt: "Benchmarks estáticos não morreram, mas já não bastam. Da prova à agência: harnesses, duração, custo, repetição e valor econômico." }
6
8
  topics: ["AI Evaluation", "Agentic AI", "Benchmarks"]
7
9
  featured: true
8
10
  host: "articles.santismm.com"
9
11
  canonicalUrl: "https://articles.santismm.com/the-stopwatch-and-the-exam/"
10
12
  links:
11
13
  - { platform: "site", lang: "en", url: "https://articles.santismm.com/the-stopwatch-and-the-exam/" }
14
+ - { platform: "site", lang: "es", url: "https://articles.santismm.com/el-cronometro-y-el-examen/" }
15
+ - { platform: "site", lang: "pt", url: "https://articles.santismm.com/o-cronometro-e-o-exame/" }
12
16
  ---