skill package

Agent Observer

Traces agent runs, tool calls, costs, permission decisions, and failure points for debugging production workflows.

ObservabilityVerifiedvv1.4.0publicAdmin

96% eval4.9 rating1,842 installs6 targets

Edit in Build

Permissions, source, and what this skill claims to do.

Contract

What this skill may access, and where it can install.

read_fileslow risk

Read trace exports and run manifests.

networkmedium risk

Fetch remote trace bundles and provider metadata.

api_keyshigh risk

Read redacted provider routing metadata.

Compatibility

CodexClaudeAntigravityOpenCodeGrokVS Code

Source

README and SKILL.md for the current version.

README

Agent Observer turns raw agent activity into readable timelines with cost, latency, tool-call, and failure annotations.

SKILL.md

---
name: agent-observer
description: agent observability and trace debugging
---

# Agent Observer

Use this skill when the user needs agent observability and trace debugging.

## Workflow
1. Confirm the user's goal and available context.
2. Inspect relevant files, docs, or input data.
3. Execute only approved actions.
4. Return a concise result with traceable evidence.

## Permissions
- read_files
- network
- api_keys

## Examples
- "Run this skill against the current repo."
- "Generate a report with cited evidence."
- "Package this result for a downstream agent."

Reception

Community feedback for this skill.

No reviews yet.

Loading studio…