From 216aa97148b922a50ffe771543e5e81e7a390800 Mon Sep 17 00:00:00 2001 From: soneeee22000 <109932809+soneeee22000@users.noreply.github.com> Date: Thu, 7 May 2026 04:12:45 +0200 Subject: [PATCH] feat(projects): add AgentProbe to featured portfolio ReAct Agent Observatory with 8-type failure taxonomy and multi-provider eval harness. Surfaces the agent-reliability work alongside existing data and backend projects. --- src/lib/data.ts | 10 ++++++++++ src/messages/de.json | 3 ++- src/messages/en.json | 3 ++- src/messages/es.json | 3 ++- src/messages/fr.json | 3 ++- src/messages/ja.json | 3 ++- src/messages/ko.json | 3 ++- src/messages/my.json | 3 ++- src/messages/pt.json | 3 ++- src/messages/th.json | 3 ++- src/messages/zh.json | 3 ++- 11 files changed, 30 insertions(+), 10 deletions(-) diff --git a/src/lib/data.ts b/src/lib/data.ts index 5e3a14c..68887d5 100644 --- a/src/lib/data.ts +++ b/src/lib/data.ts @@ -158,6 +158,16 @@ export const PROJECTS_META = [ demo: "https://greenlens-api.azurewebsites.net", gh: "https://github.com/soneeee22000/GreenLens.dev", }, + { + id: 12, + featured: true, + emoji: "\uD83D\uDD2C", + title: "AgentProbe", + tags: ["Python", "FastAPI", "Next.js 16", "PostgreSQL", "ReAct", "SSE"], + color: "#C9A96E", + demo: "#", + gh: "https://github.com/soneeee22000/AgentProbe", + }, ]; export const SKILLS: Skill[] = [ diff --git a/src/messages/de.json b/src/messages/de.json index df4d65c..227d417 100644 --- a/src/messages/de.json +++ b/src/messages/de.json @@ -88,7 +88,8 @@ "7_desc": "Real-time urban mobility analytics on TimescaleDB + PostGIS + Uber H3. Spring Boot ingestion, Kafka stream processing, Server-Sent Events for live dashboards. Smart-city back-end blueprint.", "8_desc": "Reference implementation of the BCBS 239 risk-data-aggregation lakehouse pattern on Databricks + Delta Lake + Unity Catalog + dbt-databricks. MIT-licensed, synthetic data only — banking risk-data engineering portfolio piece.", "9_desc": "Serverless middleware enforcing responsible-AI compliance on LLM applications — PII detection, bias screening, hate-speech filtering, and RAG-powered policy enforcement on top of Azure OpenAI.", - "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance." + "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance.", + "11_desc": "From-scratch ReAct Agent Observatory — observe, debug, and benchmark LLM agents with a built-in 8-type failure taxonomy (hallucinated tools, malformed actions, context overflow, goal drift, …) and multi-provider eval harness (Groq, OpenAI, Anthropic, Google, Ollama). Composite scoring (answer + tools + efficiency + reliability), Clean Architecture FastAPI + PostgreSQL backend, Next.js 16 frontend with real-time SSE streaming. 50+ benchmark cases, 81 tests." }, "skills": { "label": "04 — SKILLS", diff --git a/src/messages/en.json b/src/messages/en.json index 9efa11c..56b7db9 100644 --- a/src/messages/en.json +++ b/src/messages/en.json @@ -88,7 +88,8 @@ "7_desc": "Real-time urban mobility analytics on TimescaleDB + PostGIS + Uber H3. Spring Boot ingestion, Kafka stream processing, Server-Sent Events for live dashboards. Smart-city back-end blueprint.", "8_desc": "Reference implementation of the BCBS 239 risk-data-aggregation lakehouse pattern on Databricks + Delta Lake + Unity Catalog + dbt-databricks. MIT-licensed, synthetic data only — banking risk-data engineering portfolio piece.", "9_desc": "Serverless middleware enforcing responsible-AI compliance on LLM applications — PII detection, bias screening, hate-speech filtering, and RAG-powered policy enforcement on top of Azure OpenAI.", - "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance." + "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance.", + "11_desc": "From-scratch ReAct Agent Observatory — observe, debug, and benchmark LLM agents with a built-in 8-type failure taxonomy (hallucinated tools, malformed actions, context overflow, goal drift, …) and multi-provider eval harness (Groq, OpenAI, Anthropic, Google, Ollama). Composite scoring (answer + tools + efficiency + reliability), Clean Architecture FastAPI + PostgreSQL backend, Next.js 16 frontend with real-time SSE streaming. 50+ benchmark cases, 81 tests." }, "skills": { "label": "04 — SKILLS", diff --git a/src/messages/es.json b/src/messages/es.json index 0524042..778b8cd 100644 --- a/src/messages/es.json +++ b/src/messages/es.json @@ -88,7 +88,8 @@ "7_desc": "Real-time urban mobility analytics on TimescaleDB + PostGIS + Uber H3. Spring Boot ingestion, Kafka stream processing, Server-Sent Events for live dashboards. Smart-city back-end blueprint.", "8_desc": "Reference implementation of the BCBS 239 risk-data-aggregation lakehouse pattern on Databricks + Delta Lake + Unity Catalog + dbt-databricks. MIT-licensed, synthetic data only — banking risk-data engineering portfolio piece.", "9_desc": "Serverless middleware enforcing responsible-AI compliance on LLM applications — PII detection, bias screening, hate-speech filtering, and RAG-powered policy enforcement on top of Azure OpenAI.", - "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance." + "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance.", + "11_desc": "From-scratch ReAct Agent Observatory — observe, debug, and benchmark LLM agents with a built-in 8-type failure taxonomy (hallucinated tools, malformed actions, context overflow, goal drift, …) and multi-provider eval harness (Groq, OpenAI, Anthropic, Google, Ollama). Composite scoring (answer + tools + efficiency + reliability), Clean Architecture FastAPI + PostgreSQL backend, Next.js 16 frontend with real-time SSE streaming. 50+ benchmark cases, 81 tests." }, "skills": { "label": "04 — SKILLS", diff --git a/src/messages/fr.json b/src/messages/fr.json index 6e801b0..3d1ae61 100644 --- a/src/messages/fr.json +++ b/src/messages/fr.json @@ -88,7 +88,8 @@ "7_desc": "Real-time urban mobility analytics on TimescaleDB + PostGIS + Uber H3. Spring Boot ingestion, Kafka stream processing, Server-Sent Events for live dashboards. Smart-city back-end blueprint.", "8_desc": "Reference implementation of the BCBS 239 risk-data-aggregation lakehouse pattern on Databricks + Delta Lake + Unity Catalog + dbt-databricks. MIT-licensed, synthetic data only — banking risk-data engineering portfolio piece.", "9_desc": "Serverless middleware enforcing responsible-AI compliance on LLM applications — PII detection, bias screening, hate-speech filtering, and RAG-powered policy enforcement on top of Azure OpenAI.", - "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance." + "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance.", + "11_desc": "From-scratch ReAct Agent Observatory — observe, debug, and benchmark LLM agents with a built-in 8-type failure taxonomy (hallucinated tools, malformed actions, context overflow, goal drift, …) and multi-provider eval harness (Groq, OpenAI, Anthropic, Google, Ollama). Composite scoring (answer + tools + efficiency + reliability), Clean Architecture FastAPI + PostgreSQL backend, Next.js 16 frontend with real-time SSE streaming. 50+ benchmark cases, 81 tests." }, "skills": { "label": "04 — SKILLS", diff --git a/src/messages/ja.json b/src/messages/ja.json index 43946a7..3cdfbcf 100644 --- a/src/messages/ja.json +++ b/src/messages/ja.json @@ -88,7 +88,8 @@ "7_desc": "Real-time urban mobility analytics on TimescaleDB + PostGIS + Uber H3. Spring Boot ingestion, Kafka stream processing, Server-Sent Events for live dashboards. Smart-city back-end blueprint.", "8_desc": "Reference implementation of the BCBS 239 risk-data-aggregation lakehouse pattern on Databricks + Delta Lake + Unity Catalog + dbt-databricks. MIT-licensed, synthetic data only — banking risk-data engineering portfolio piece.", "9_desc": "Serverless middleware enforcing responsible-AI compliance on LLM applications — PII detection, bias screening, hate-speech filtering, and RAG-powered policy enforcement on top of Azure OpenAI.", - "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance." + "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance.", + "11_desc": "From-scratch ReAct Agent Observatory — observe, debug, and benchmark LLM agents with a built-in 8-type failure taxonomy (hallucinated tools, malformed actions, context overflow, goal drift, …) and multi-provider eval harness (Groq, OpenAI, Anthropic, Google, Ollama). Composite scoring (answer + tools + efficiency + reliability), Clean Architecture FastAPI + PostgreSQL backend, Next.js 16 frontend with real-time SSE streaming. 50+ benchmark cases, 81 tests." }, "skills": { "label": "04 — SKILLS", diff --git a/src/messages/ko.json b/src/messages/ko.json index e403dfc..cfbb4bb 100644 --- a/src/messages/ko.json +++ b/src/messages/ko.json @@ -88,7 +88,8 @@ "7_desc": "Real-time urban mobility analytics on TimescaleDB + PostGIS + Uber H3. Spring Boot ingestion, Kafka stream processing, Server-Sent Events for live dashboards. Smart-city back-end blueprint.", "8_desc": "Reference implementation of the BCBS 239 risk-data-aggregation lakehouse pattern on Databricks + Delta Lake + Unity Catalog + dbt-databricks. MIT-licensed, synthetic data only — banking risk-data engineering portfolio piece.", "9_desc": "Serverless middleware enforcing responsible-AI compliance on LLM applications — PII detection, bias screening, hate-speech filtering, and RAG-powered policy enforcement on top of Azure OpenAI.", - "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance." + "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance.", + "11_desc": "From-scratch ReAct Agent Observatory — observe, debug, and benchmark LLM agents with a built-in 8-type failure taxonomy (hallucinated tools, malformed actions, context overflow, goal drift, …) and multi-provider eval harness (Groq, OpenAI, Anthropic, Google, Ollama). Composite scoring (answer + tools + efficiency + reliability), Clean Architecture FastAPI + PostgreSQL backend, Next.js 16 frontend with real-time SSE streaming. 50+ benchmark cases, 81 tests." }, "skills": { "label": "04 — SKILLS", diff --git a/src/messages/my.json b/src/messages/my.json index bb05d4c..ae372cf 100644 --- a/src/messages/my.json +++ b/src/messages/my.json @@ -88,7 +88,8 @@ "7_desc": "Real-time urban mobility analytics on TimescaleDB + PostGIS + Uber H3. Spring Boot ingestion, Kafka stream processing, Server-Sent Events for live dashboards. Smart-city back-end blueprint.", "8_desc": "Reference implementation of the BCBS 239 risk-data-aggregation lakehouse pattern on Databricks + Delta Lake + Unity Catalog + dbt-databricks. MIT-licensed, synthetic data only — banking risk-data engineering portfolio piece.", "9_desc": "Serverless middleware enforcing responsible-AI compliance on LLM applications — PII detection, bias screening, hate-speech filtering, and RAG-powered policy enforcement on top of Azure OpenAI.", - "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance." + "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance.", + "11_desc": "From-scratch ReAct Agent Observatory — observe, debug, and benchmark LLM agents with a built-in 8-type failure taxonomy (hallucinated tools, malformed actions, context overflow, goal drift, …) and multi-provider eval harness (Groq, OpenAI, Anthropic, Google, Ollama). Composite scoring (answer + tools + efficiency + reliability), Clean Architecture FastAPI + PostgreSQL backend, Next.js 16 frontend with real-time SSE streaming. 50+ benchmark cases, 81 tests." }, "skills": { "label": "04 — SKILLS", diff --git a/src/messages/pt.json b/src/messages/pt.json index 4a7d211..115af54 100644 --- a/src/messages/pt.json +++ b/src/messages/pt.json @@ -88,7 +88,8 @@ "7_desc": "Real-time urban mobility analytics on TimescaleDB + PostGIS + Uber H3. Spring Boot ingestion, Kafka stream processing, Server-Sent Events for live dashboards. Smart-city back-end blueprint.", "8_desc": "Reference implementation of the BCBS 239 risk-data-aggregation lakehouse pattern on Databricks + Delta Lake + Unity Catalog + dbt-databricks. MIT-licensed, synthetic data only — banking risk-data engineering portfolio piece.", "9_desc": "Serverless middleware enforcing responsible-AI compliance on LLM applications — PII detection, bias screening, hate-speech filtering, and RAG-powered policy enforcement on top of Azure OpenAI.", - "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance." + "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance.", + "11_desc": "From-scratch ReAct Agent Observatory — observe, debug, and benchmark LLM agents with a built-in 8-type failure taxonomy (hallucinated tools, malformed actions, context overflow, goal drift, …) and multi-provider eval harness (Groq, OpenAI, Anthropic, Google, Ollama). Composite scoring (answer + tools + efficiency + reliability), Clean Architecture FastAPI + PostgreSQL backend, Next.js 16 frontend with real-time SSE streaming. 50+ benchmark cases, 81 tests." }, "skills": { "label": "04 — SKILLS", diff --git a/src/messages/th.json b/src/messages/th.json index 21abae2..9b7902c 100644 --- a/src/messages/th.json +++ b/src/messages/th.json @@ -88,7 +88,8 @@ "7_desc": "Real-time urban mobility analytics on TimescaleDB + PostGIS + Uber H3. Spring Boot ingestion, Kafka stream processing, Server-Sent Events for live dashboards. Smart-city back-end blueprint.", "8_desc": "Reference implementation of the BCBS 239 risk-data-aggregation lakehouse pattern on Databricks + Delta Lake + Unity Catalog + dbt-databricks. MIT-licensed, synthetic data only — banking risk-data engineering portfolio piece.", "9_desc": "Serverless middleware enforcing responsible-AI compliance on LLM applications — PII detection, bias screening, hate-speech filtering, and RAG-powered policy enforcement on top of Azure OpenAI.", - "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance." + "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance.", + "11_desc": "From-scratch ReAct Agent Observatory — observe, debug, and benchmark LLM agents with a built-in 8-type failure taxonomy (hallucinated tools, malformed actions, context overflow, goal drift, …) and multi-provider eval harness (Groq, OpenAI, Anthropic, Google, Ollama). Composite scoring (answer + tools + efficiency + reliability), Clean Architecture FastAPI + PostgreSQL backend, Next.js 16 frontend with real-time SSE streaming. 50+ benchmark cases, 81 tests." }, "skills": { "label": "04 — SKILLS", diff --git a/src/messages/zh.json b/src/messages/zh.json index 46ea86a..ba2486d 100644 --- a/src/messages/zh.json +++ b/src/messages/zh.json @@ -88,7 +88,8 @@ "7_desc": "Real-time urban mobility analytics on TimescaleDB + PostGIS + Uber H3. Spring Boot ingestion, Kafka stream processing, Server-Sent Events for live dashboards. Smart-city back-end blueprint.", "8_desc": "Reference implementation of the BCBS 239 risk-data-aggregation lakehouse pattern on Databricks + Delta Lake + Unity Catalog + dbt-databricks. MIT-licensed, synthetic data only — banking risk-data engineering portfolio piece.", "9_desc": "Serverless middleware enforcing responsible-AI compliance on LLM applications — PII detection, bias screening, hate-speech filtering, and RAG-powered policy enforcement on top of Azure OpenAI.", - "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance." + "10_desc": "Cloud Carbon Intelligence platform estimating CO2e emissions from Azure infrastructure with AI-powered reduction recommendations. Clean Architecture, Azure AI Search semantic factor lookup, 88 automated tests. Built for EU CSRD Scope 3 compliance.", + "11_desc": "From-scratch ReAct Agent Observatory — observe, debug, and benchmark LLM agents with a built-in 8-type failure taxonomy (hallucinated tools, malformed actions, context overflow, goal drift, …) and multi-provider eval harness (Groq, OpenAI, Anthropic, Google, Ollama). Composite scoring (answer + tools + efficiency + reliability), Clean Architecture FastAPI + PostgreSQL backend, Next.js 16 frontend with real-time SSE streaming. 50+ benchmark cases, 81 tests." }, "skills": { "label": "04 — SKILLS",