Show HN: WeaveScope – Elixir native observability for AI agents

WeaveScope, an Elixir-native observability tool for AI agents, launched in early access, offering tracing and monitoring for agent runs with a free tier of up to 100,000 accepted events per month. The tool, which requires one Elixir dependency, enables inspection of model calls, tool calls, and spans, with pricing starting at €0 per month for the Free plan and €29 per month for Pro.

No prior 43,040 Analytics unit WeaveScope is in early access — tracing and monitoring for Elixir agents Trace the model or tool call behind failed, slow, or expensive Elixir agent runs. Free up to 100K accepted events / month · no credit card · one Elixir dependency Inspect model calls, tool calls, and spans across instrumented agent runs. How it works integrate | Trace ID | Name | User | Model | Spans | Tags | Tokens | Cost | Total time | When | | |---|---|---|---|---|---|---|---|---|---|---| | 019fd6f2...0ea0 | x signal desk.post summary | anonymous | zai:glm-5.2 | 9 | 5,646 | $0.0126 | 21.79 s | 2026-08-06 12:01 | || | 019fd6f2...9bcd | x signal desk.post summary | anonymous | zai:glm-5.2 | 9 | 5,426 | $0.0120 | 17.71 s | 2026-08-06 12:01 | || | 019fd6d7...60c2 | x signal desk.post summary | anonymous | zai:glm-5.2 | 9 | 4,365 | $0.0091 | 13.61 s | 2026-08-06 11:31 | || | 019fd684...4fab | x signal desk.post summary | anonymous | zai:glm-5.2 | 9 | 7,029 | $0.0173 | 32.78 s | 2026-08-06 10:01 | || | 019fd684...fe35 | x signal desk.post summary | anonymous | zai:glm-5.2 | 9 | 4,528 | $0.0093 | 18.17 s | 2026-08-06 10:01 | || | 019fd64d...54d5 | x signal desk.post summary | anonymous | zai:glm-5.2 | 9 | 7,117 | $0.0179 | 40.66 s | 2026-08-06 09:01 | || | 019fd64d...93d5 | x signal desk.post summary | anonymous | zai:glm-5.2 | 9 | 4,601 | $0.0100 | 23.68 s | 2026-08-06 09:01 | || | 019fd632...9e98 | x signal desk.post summary | anonymous | zai:glm-5.2 | 9 | 5,593 | $0.0120 | 30.45 s | 2026-08-06 08:31 | No tool calls were recorded for this observation tree. No tool calls were recorded for this observation tree. Built for how agents actually run Tracing BeamWeaver connects instrumented model calls, tool calls, and spans across an agent run. Jump from a failed step to its inputs, outputs, latency, and cost. Monitoring Follow throughput, latency, errors, token usage, and cost as your agents run. Give your team a shared view of the work behind every outcome. Alerts Coming soon Alerting is next. We are building it on top of the tracing and monitoring data your team already has—not as a separate black box. One dependency. Trace context. Add the dependency and configure one endpoint. BeamWeaver exports instrumented agent runs, model calls, and tool calls in the background. defp deps do {:beam weaver, "~ 0.1.15"} end config :beam weaver, openai: api key: System.fetch env "OPENAI API KEY" , weave scope: api key: System.fetch env "WEAVESCOPE API KEY" , endpoint: "https://app.weavescope.com" defmodule Support.Agent do alias BeamWeaver.Agent alias BeamWeaver.Core.Message def run query, context \\ %{} do model = BeamWeaver.Models.init chat model "openai:gpt-5.4-mini", temperature: 0.2, timeout: 30 000 {:ok, agent} = Agent.build name: "support.reply", model: model, system prompt: "Answer support questions clearly." Agent.invoke agent, %{messages: Message.user query }, trace: name: "support.reply", user id: context :user id , execution mode: "support reply", fields: %{ticket id: context :ticket id , account id: context :account id } end end Pricing Start free. Usage counts accepted observation events; a trace can contain multiple events. Free Community support €0/ month Pro In-app support €29/ month Scale Priority support €99/ month Enterprise Dedicated support Custom Ship with context Start with the traces and operating signals you need to make the next decision.