Skip to content
Back to skills

Review Agent Observability

ASecurity

This skill reviews agent projects for observability completeness — tracing, metrics, structured logging, and trace propagation. Use this skill when the user asks to "review agent observability", "audit agent tracing", "check agent metrics", or similar.

  • 12 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added September 5, 2026
ai-agentspythondocker

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned September 5, 2026

npx -y skills add diagrid-labs/dapr-skills --skill review-agent-observability --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Review Agent Observability?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Review Agent Observability
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/diagrid-labs-review-agent-observability/badge)](https://www.skillsdirectory.com/skills/diagrid-labs-review-agent-observability)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: review-agent-observability
description: This skill reviews agent projects for observability completeness — tracing, metrics, structured logging, and trace propagation. Use this skill when the user asks to "review agent observability", "audit agent tracing", "check agent metrics", or similar.
allowed-tools:
  - Read
  - Grep
  - Glob
---

# Review Agents — Observability

## Overview

Scans native `dapr-agents` projects for observability completeness: Dapr tracing Configuration present, sampling rate non-zero, structured logging in use, explicit log level, trace-id propagation on outbound tool calls, and README guidance for trace / metrics UIs. Framework-wrapper projects (LangGraph / CrewAI / OpenAI Agents / etc.) receive a single `info` finding — those frameworks ship their own observability pipelines.

## Execution Order

You MUST follow these phases in strict order.

1. **Resolve scope** — Read [`../shared/review-scope-prompt.md`](../shared/review-scope-prompt.md) and follow it to set `scope_root`.
2. **Detect target** — Read [`../shared/review-detect-target-agent.md`](../shared/review-detect-target-agent.md). If `flavor == framework-wrapper`, emit one `info` finding explaining that observability is delegated to the wrapped framework and stop.
3. **Load checklist** — Based on `language`, read exactly one of:
   - `python` → [`../shared/review-agent-observability-python.md`](../shared/review-agent-observability-python.md)
   - `dotnet` → [`../shared/review-agent-observability-dotnet.md`](../shared/review-agent-observability-dotnet.md) (stub — emits a single `DAG-OBS-000` info finding)
4. **Scan** — Apply every rule. `dapr.yaml`, `resources/tracing.yaml`, tool file contents, and `README.md` are all potentially in scope.
5. **Report** — Format findings using [`../shared/review-report-format.md`](../shared/review-report-format.md).
6. **Show final message** — Emit the report from step 5 with a `## Next steps` block.

## Prerequisites

- Read access to the project directory.

## Allowed tools

`Read`, `Grep`, `Glob` only.

## Rules

- [Python rules](../shared/review-agent-observability-python.md) — `DAG-OBS-001` … `DAG-OBS-008`
- [.NET rules](../shared/review-agent-observability-dotnet.md) — `DAG-OBS-000` (stub; full checklist TBD)

## Show final message

The last thing you emit MUST be the report from step 5. `## Next steps` should suggest:

- Run `review-agent-tools`, `review-agent-memory`, and `review-agent-orchestration` if not yet run.
- Re-run this skill after fixing critical findings.
- Optionally run the observability stack locally (`docker compose -f docker-compose.observability.yaml up -d`) and verify traces appear in Zipkin.

Do not add any text after the report.

Files in this skill

  • REFERENCE.md2.2 KB
  • SKILL.md2.7 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.