Osservabilità LLM

Classifica degli strumenti

Classifica per crescita misurata e normalizzata tra le fonti. I valori grezzi mantengono la propria finestra; una variazione richiede almeno due rilevazioni in 7 giorni.

1. 1

   astragraph

   Agente

   DormienteUltimo commit 5 mesi fa · 2026-03-16

   Policy-enforced observability and fail-closed guardrails for MCP/A2A multi-agent systems.

   Rust

   • a2a
   • ai-agents
   • llm-security
   • mcp
   • multi-agent

   githubcreato 6 mesi facrescita misurataApri fonte ↗

   26stelle GitHubstabile
2. 2

   otel-agent-provenance

   Agente

   DormienteUltimo commit 5 mesi fa · 2026-03-17

   OpenTelemetry semantic conventions and instrumentation for agent provenance, derivation lineage, and acceptance criteria evaluation. Fills the Microsoft AI stack observability gap.

   Python

   • a2a-protocol
   • agentic-ai
   • ai-agents
   • observability
   • opentelemetry

   githubcreato 5 mesi facrescita misurataApri fonte ↗

   0stelle GitHubstabile
3. 3

   sentry-instrumentation

   Skill

   AttivoUltimo commit 2 mesi fa · 2026-06-19

   Sentry instrumentation skill for system-behavior tracking

   ShellLicenza: MIT

   • agent-skills
   • claude-skills
   • codex-skill
   • instrumentation
   • monitoring

   githubcreato 4 mesi facrescita misurataApri fonte ↗

   Installa git clone https://github.com/tortastudios/sentry-instrumentation ~/.claude/skills/sentry-instrumentation

   24stelle GitHubstabile
4. 4

   bar-observatory

   MCP

   AttivoUltimo commit 6 giorni fa · 2026-08-27

   Deterministic, local-only audit reports for Claude Code AI agent sessions. Rust + SQLite. Zero network calls. MCP server for agents.

   HTMLLicenza: MIT

   • agent-monitoring
   • agentic-ai
   • ai-agents
   • ai-engineering
   • audit

   githubcreato 21 giorni fanuovocrescita misurataApri fonte ↗

   22stelle GitHub+2 (+10.0 %)
5. 5

   opik-claude-code-plugin

   MCP

   AttivoUltimo commit 34 giorni fa · 2026-07-30

   Log Claude Code sessions to Opik, the open-source LLM observability and evaluation platform, built by Comet. Tracing, evaluation, and skills for observable AI applications.

   GoLicenza: Apache-2.0

   • ai-agent
   • claude-ai
   • claude-code-plugin
   • llm-observation
   • mcp-server

   githubcreato 6 mesi facrescita misurataApri fonte ↗

   22stelle GitHub+1 (+4.8 %)
6. 6

   AI Guardian

   MCP

   AttivoUltimo commit 8 giorni fa · 2026-08-26

   Governed local-LLM observability: model policy, prompt scanner, capture proxy, 21 tools.

   PythonLicenza: MIT

   mcpcreato 53 giorni facrescita misurataApri fonte ↗

   Installa claude mcp add ai-guardian -- uvx ai-guardian-aiops

   0stelle GitHubstabile
7. 7

   untell

   Altro

   AttivoUltimo commit 1 giorno fa · 2026-09-02

   AI-detector auditing toolkit: measures how often a detector flags genuine human writing, how stable its verdict is across seeds, and whether that verdict survives meaning-preserving edits. Detector-in-the-loop measurement harness. Claude…

   PythonLicenza: MIT

   • ai-detection
   • ai-detector
   • benchmark
   • claude-code
   • claude-code-skill

   githubcreato 2 mesi facrescita misurataApri fonte ↗

   18stelle GitHubstabile
8. 8

   rashomon

   Skill

   AttivoUltimo commit 3 giorni fa · 2026-08-30

   Measure prompt and skill improvements with blind A/B comparison.

   PythonLicenza: MIT

   • agent-skills
   • ai-tools
   • claude-code
   • claude-code-plugin
   • claude-code-skill

   githubcreato 7 mesi facrescita misurataApri fonte ↗

   Installa /plugin marketplace add shinpr/rashomon

   18stelle GitHubstabile
9. 9

   parallel-adversarial-review

   Skill

   AttivoUltimo commit 2 mesi fa · 2026-06-06

   Two Claude Code skills for adversarial code review (single-model PAR and multi-model MMAR with cross-critique to catch hallucinations), plus a fixture-based eval suite.

   Python

   • ai-agents
   • claude-code-plugin
   • code-review
   • evals
   • python

   githubcreato 3 mesi facrescita misurataApri fonte ↗

   Installa /plugin marketplace add prime-radiant-inc/parallel-adversarial-review

   17stelle GitHubstabile
10. 10

    simple-output-styles

    Skill

    AttivoUltimo commit 1 giorno fa · 2026-09-02

    Make Claude write clearly, for everyone.

    PythonLicenza: 0BSD

    • claude-code
    • claude-code-plugin
    • controlled-language
    • llm-evaluation
    • output-style

    githubcreato 33 giorni facrescita misurataApri fonte ↗

    Installa /plugin marketplace add stefanobaghino/simple-output-styles

    16stelle GitHubstabile
11. 11

    agentops

    Agente

    DormienteUltimo commit 4 mesi fa · 2026-04-13

    AgentOps: Multi-agent infrastructure remediation platform. A2A protocol, agent coordination, HITL approval, auto-rollback.

    PythonLicenza: MIT

    • a2a-protocol
    • ai
    • automation
    • infrastructure
    • llm

    githubcreato 6 mesi facrescita misurataApri fonte ↗

    0stelle GitHubstabile
12. 12

    adl-cli

    Altro

    AttivoUltimo commit 4 giorni fa · 2026-08-29

    A command-line tool to scaffold and manage enterprise-ready AI Agents powered by the A2A (Agent-to-Agent) protocol

    GoLicenza: Apache-2.0

    • a2a
    • agents
    • ai
    • containerized
    • docker

    githubcreato 13 mesi facrescita misurataApri fonte ↗

    14stelle GitHub+1 (+7.7 %)
13. 13

    agentic-engineering-marketplace

    Skill

    AttivoUltimo commit 36 giorni fa · 2026-07-29

    Claude Code plugin marketplace — agentic-engineering (spec-driven shape→decide→execute→measure→eval with adversarial review) + github-keeper (audit/elevate READMEs and make a repo open-source-ready).

    PythonLicenza: MIT

    • adr
    • agentic-engineering
    • ai-agents
    • anthropic
    • claude-code

    githubcreato 2 mesi facrescita misurataApri fonte ↗

    Installa /plugin marketplace add GiustoPiedimonte/agentic-engineering-marketplace

    13stelle GitHubstabile
14. 14

    headsup

    Skill

    AttivoUltimo commit 15 giorni fa · 2026-08-18

    Glanceable Claude Code and Codex state in your terminal tabs: white=idle, blue=working, orange=waiting. Multi-terminal (iTerm2, WezTerm, AI Power Term), a live status bar with Anthropic usage limits, /sfl and /nil window save-and-restore,…

    ShellLicenza: MIT

    • ai-workflow-observability
    • claude-code
    • claude-code-hooks
    • claude-code-skill
    • claude-code-skills

    githubcreato 3 mesi facrescita misurataApri fonte ↗

    Installa git clone https://github.com/wasulajr/headsup ~/.claude/skills/headsup

    1stelle GitHubstabile
15. 15

    eval-layer

    Altro

    DormienteUltimo commit 4 mesi fa · 2026-04-29

    A Claude Code skill that adds a rubric-based eval layer to any agent project. Framework-agnostic — generates rubric, test cases, judge prompt, and harness. Returns a weighted score plus a judge-leniency signal.

    HTML

    • agent-evaluation
    • anthropic
    • claude-code
    • claude-code-skill
    • evals

    githubcreato 4 mesi facrescita misurataApri fonte ↗

    13stelle GitHubstabile
16. 16

    io.github.spanlens/mcp-server

    MCP

    AttivoUltimo commit 10 giorni fa · 2026-08-24

    Query Spanlens LLM observability from Cursor, Claude Desktop, or Continue via MCP.

    ↓ 32 download/settimanaTypeScriptLicenza: MIT

    mcpcreato 4 mesi facrescita misurataApri fonte ↗

    Installa claude mcp add mcp-server -- npx @spanlens/mcp-server

    12stelle GitHubstabile
17. 17

    AgentStack

    Agente

    AttivoUltimo commit 2 giorni fa · 2026-09-01

    Provide clear documentation for AgentStack’s MCP protocol, plugins, and ecosystem API with usage examples and tool references.

    • a2a
    • anthropic
    • claude-code
    • crewai
    • deep-learning

    githubcreato 5 mesi facrescita misurataApri fonte ↗

    0stelle GitHubstabile
18. 18

    ai-dev-stack

    Skill

    AttivoUltimo commit 2 mesi fa · 2026-06-29

    Production-grade AI coding rules for Cursor and Claude Code. 15 rules + 9 doc templates + skills + agents + MCP setup. Drop into any project.

    PythonLicenza: MIT

    • ai-agents
    • ai-coding
    • ai-development
    • ai-engineer
    • automation

    githubcreato 5 mesi facrescita misurataApri fonte ↗

    Installa git clone https://github.com/aiagentwithdhruv/ai-dev-stack ~/.claude/skills/ai-dev-stack

    10stelle GitHub+1 (+11.1 %)
19. 19

    shokunin-review

    Altro

    AttivoUltimo commit 2 mesi fa · 2026-06-25

    Terminal-first validation harness for reviewing PRDs, RFCs, strategy docs, and experiment plans.

    TypeScript

    • ai-agents
    • claude-code
    • claude-code-skills
    • cli
    • evals

    githubcreato 2 mesi facrescita misurataApri fonte ↗

    1stelle GitHubstabile
20. 20

    skill-graveyard

    Skill

    AttivoUltimo commit 55 giorni fa · 2026-07-10

    Audit which Claude Code skills you actually use — surface dead installs and hallucinated invocations from your session logs.

    TypeScriptLicenza: MIT

    • anthropic
    • audit
    • claude
    • claude-code
    • claude-code-plugin

    githubcreato 4 mesi facrescita misurataApri fonte ↗

    Installa /plugin marketplace add sfrangulov/skill-graveyard

    10stelle GitHub+1 (+11.1 %)
21. 21

    bakeoff

    Skill

    AttivoUltimo commit 55 giorni fa · 2026-07-09

    Turn one decision into a judged tournament of solutions, then pick the best — a Claude Code skill that generates candidates, auto-derives the rubric, judges independently, and returns a defensible winner.

    JavaScriptLicenza: MIT

    • ai-agents
    • anthropic
    • claude-code
    • claude-code-plugin
    • claude-skill

    githubcreato 58 giorni facrescita misurataApri fonte ↗

    Installa /plugin marketplace add CoriChui/bakeoff

    10stelle GitHubstabile
22. 22

    galdor

    Altro

    AttivoUltimo commit 10 giorni fa · 2026-08-24

    A Go-native framework for LLM agents, with OpenTelemetry observability built in.

    GoLicenza: Apache-2.0

    • a2a
    • agents
    • ai
    • anthropic
    • bedrock

    githubcreato 3 mesi facrescita misurataApri fonte ↗

    10stelle GitHubstabile
23. 23

    anchor

    Skill

    AttivoUltimo commit 14 giorni fa · 2026-08-20

    Anchor — the production-grade AGENTS.md template for AI coding agents. 51 sections of battle-tested rules. Keep your agents grounded.

    Licenza: MIT

    • agent-instructions
    • agents-md
    • ai-agent-instructions
    • ai-coding
    • ai-coding-agent

    githubcreato 3 mesi facrescita misurataApri fonte ↗

    Installa git clone https://github.com/peva3/anchor ~/.claude/skills/anchor

    8stelle GitHubstabile
24. 24

    adaptive-harness

    Skill

    DormienteUltimo commit 5 mesi fa · 2026-03-16

    A self-improving harness router for Claude Code.

    PythonLicenza: MIT

    • a-b-testing
    • ai-agent
    • ai-orchestration
    • ai-routing
    • ai-workflow

    githubcreato 5 mesi facrescita misurataApri fonte ↗

    Installa /plugin marketplace add SeongwoongCho/adaptive-harness

    8stelle GitHubstabile
25. 25

    tulip-agents

    Agente

    AttivoUltimo commit 7 giorni fa · 2026-08-26

    The agent framework where the model never holds the trigger — every consequential action clears your policy first, waits for a human when it matters, and lands on a record you can verify. Build on it, or put it around the agent you already…

    PythonLicenza: Apache-2.0

    • a2a
    • agent-framework
    • ai-agents
    • ai-governance
    • ai-safety

    githubcreato 2 mesi facrescita misurataApri fonte ↗

    1stelle GitHubstabile