Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Inspect

ASecurity

Inspect a running service's runtime state — exercise API endpoints, verify internal state via backoffice, and validate the UI via browser. Spec-driven.

7 stars
0 votes
0 copies
0 views
Added 9/20/2026
developmentjavascriptjavaapi

Works with

cliapi

Security Analysis

A100/100

Scanned 9/20/2026

Install to Claude Code

$npx -y skills add akka/ai-marketplace --skill inspect --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Inspect?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Inspect
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/akka-inspect/badge)](https://www.skillsdirectory.com/skills/akka-inspect)

More formats (shields.io, HTML) on the badges page.

Download with Pro
Files
SKILL.md
---
name: inspect
description: "Inspect a running service's runtime state — exercise API endpoints, verify internal state via backoffice, and validate the UI via browser. Spec-driven."
---

## User Input

You **MUST** consider the user input before proceeding (if not empty).

## Purpose

This command inspects a **running** service at runtime. It exercises API
endpoints, verifies internal state through backoffice tools, and validates the
UI through browser tools — all driven by the feature specification. No
compilation or unit tests happen here; use `/akka:build` for that.

## Outline

1. **Pre-flight checks**:
   - Call `akka_local_status` to verify the service is running. If not running,
     **stop** and tell the user: *"No service is running locally. Run
     `/akka:build` first to compile, test, and start the service."*
   - Call `akka_sdd_list_specs` to find the target feature. If the user
     specified a feature name in `the user's request`, match it. Otherwise, if there is
     exactly one feature, use it. If multiple features exist, ask the user which
     to inspect.
   - If no spec exists, **stop** and tell the user: *"No feature specification
     found. Run `/akka:specify` first to create one, or provide a feature name."*
   - Read the spec.md for the target feature. Extract:
     - API endpoints (paths, methods, expected request/response shapes)
     - Entities and their expected events/state
     - Workflows and expected step sequences
     - Views and expected query results
     - Agents and expected behavior
     - Acceptance criteria to verify

2. **Discover components**: Call `akka_backoffice_list_components` with
   `local=true` to list all components registered in the running service.
   - Compare against what the spec defines — flag any components that are
     missing from the running service or present but not in the spec
   - Report the component inventory: entities, views, workflows, agents,
     endpoints

3. **Exercise API endpoints**: For each endpoint defined in the spec, use
   `akka_local_request` to send requests and verify responses:
   - **Happy path**: Send valid requests matching the spec's examples or
     acceptance criteria. Verify HTTP status codes and response payloads.
   - **Command operations**: POST/PUT/DELETE commands that mutate state.
     Execute these in the order that makes sense (e.g. create before update).
   - **Query operations**: GET requests to read state back. Verify the
     response matches what the commands should have produced.
   - **Error cases**: If the spec defines error scenarios (invalid input,
     not-found, conflict), test those too. Verify appropriate error responses.
   - Record all request/response pairs for the report.

4. **Verify internal state**: After exercising the API, use backoffice tools
   with `local=true` to verify the service internals are consistent:
   - **Entity state**: `akka_backoffice_get_entity_state` — verify entity
     state matches what the API commands should have produced
   - **Event journal**: `akka_backoffice_list_events` — verify event-sourced
     entities have the expected events in the correct order
   - **Workflows**: `akka_backoffice_get_workflow` — verify workflow steps
     executed with expected status and state transitions
   - **Views**: `akka_backoffice_query_view` — verify view projections return
     expected data after entity state changes
   - **Agents**: `akka_backoffice_list_agent_interactions` — review agent
     tool calls, model responses, and guardrail behavior
   - **Timers**: `akka_backoffice_list_timers` — verify expected timers are
     registered
   - Use `akka_backoffice_describe_view` to understand view schemas before
     querying
   - Use pagination (`page_size`, `page_token`) for large result sets

5. **Browser inspection** (only if the spec defines a web UI served by the
   service). Skip this step entirely if the service only exposes API endpoints
   — use `akka_local_request` in step 3 for those.
   - `akka_browser_navigate` to open the service's UI endpoint
     (e.g. `http://localhost:<port>/ui` — use the port and path from the spec
     or from `akka_local_status`)
   - `akka_browser_screenshot` to capture the visual state
   - `akka_browser_snapshot` to verify text content
   - `akka_browser_click` and `akka_browser_fill` to interact with forms and
     buttons as described in the spec
   - `akka_browser_eval` for advanced JavaScript inspection if needed
   - Cross-reference: verify that what the UI shows matches entity/view state
     from step 4. For example, if a view query returns 3 items, the UI should
     render 3 items.
   - `akka_browser_close` when done to release browser resources

6. **Report**: Summarize inspection findings organized by the spec's acceptance
   criteria:
   - For each acceptance criterion: **pass** or **fail** with evidence
     (request/response payloads, entity state snapshots, event journals,
     screenshots)
   - Component inventory: all registered vs expected
   - API endpoint results: status codes, response validation
   - Internal state consistency: entity state, events, workflow steps, views
   - Browser findings: UI screenshots, interaction results
   - **Issues found**: list anything that doesn't match the spec, with
     severity (blocker / warning / observation)
   - **Next steps**: suggest `/akka:implement` for fixes, `/akka:review` for
     code review, or `/akka:deploy` if everything passes

## Error Handling

- If the service crashes during inspection (endpoint returns connection refused
  after previously working), check `akka_local_logs` with `source: "service"`
  for the crash reason. Report it and suggest `/akka:build` to restart.
- If a backoffice tool returns an error, report the specific error and continue
  inspecting other components — don't stop the whole inspection for one failure.
- If the browser fails to launch (Chrome not found), skip browser inspection
  and note it in the report. The backoffice and API inspection results are
  still valuable.

## Key Rules

- This is INSPECTION ONLY — do not modify code, configuration, or service state
  beyond what the API calls naturally produce
- Always use `local=true` on all backoffice tool calls
- Exercise endpoints in a logical order (create before read/update/delete)
- Report failures with evidence, not just "it failed"
- If the spec is vague about expected behavior, state what you observed and
  ask the user whether it matches their intent

## Done When

- [ ] `akka_local_status` confirmed the service is running (or the command stopped and told the user to run `/akka:build` first).
- [ ] `akka_sdd_list_specs` resolved a single target feature — matched by `the user's request`, auto-selected when only one exists, or explicitly chosen by the user — and its `spec.md` was read (or the command stopped and told the user to run `/akka:specify`).
- [ ] The spec was parsed for API endpoints, entities, workflows, views, agents, and acceptance criteria before any tool calls were made.
- [ ] `akka_backoffice_list_components` (`local=true`) produced a component inventory; any component present in the spec but missing at runtime — or present at runtime but not in the spec — was flagged.
- [ ] Every API endpoint defined in the spec was exercised via `akka_local_request` in a logical order (create before read/update/delete), covering happy-path, command, query, and any spec-defined error cases; each request/response pair was recorded for the report.
- [ ] Internal state was verified against the API's effects using backoffice tools with `local=true`: entity state (`akka_backoffice_get_entity_state`), event journals (`akka_backoffice_list_events`), workflow steps (`akka_backoffice_get_workflow`), view projections (`akka_backoffice_query_view`), agent interactions (`akka_backoffice_list_agent_interactions`), and timers (`akka_backoffice_list_timers`) — with pagination when needed.
- [ ] Browser inspection ran only if the spec defines a web UI; when it ran, `akka_browser_screenshot`, `akka_browser_snapshot`, and interaction tools captured the UI state and cross-referenced it with view/entity state, and `akka_browser_close` was called at the end. When the spec has no UI, this step was explicitly skipped.
- [ ] NO code, configuration, or service state was modified beyond what the exercised API calls naturally produced (this is inspection only).
- [ ] Every backoffice call used `local=true`; runtime crashes were reported with a link to `akka_local_logs` output rather than silently swallowed.
- [ ] The final report organizes findings by the spec's acceptance criteria (each pass/fail with evidence), lists the component inventory, API endpoint results, internal-state consistency, browser findings, issues (blocker / warning / observation), and a next-step recommendation (`/akka:implement`, `/akka:review`, or `/akka:deploy`).

Attribution

akkaakka
View sourceMore from akka →
SSkills DirectorySkills Directory

Know which skills are safe — weekly.

Best new skills + every skill we flagged as malicious. From the team that scanned 103,619.

Join free

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 (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Know which skills are safe — weekly.

Best new skills + every skill we flagged as malicious. From the team that scanned 103,619.

Join free

Related Skills

Browser Extension Developer

Use this skill when developing or maintaining browser extension code in the `browser/` directory, including Chrome/Firefox/Edge compatibility, content scripts, background scripts, or i18n updates.

284072 votes

Seo Optimizer

SEO optimization with keyword analysis, readability assessment, technical validation, content quality. Use for search rankings, blog posts, content audits, or encountering keyword density, readability scores, meta tags, schema markup errors.

2192 votes

Google Official Seo Guide

Official Google SEO guide covering search optimization, best practices, Search Console, crawling, indexing, and improving website search visibility based on official Google documentation

1862 votes

Tanstack Start

Build a full-stack TanStack Start app on Cloudflare Workers from scratch — SSR, file-based routing, server functions, D1+Drizzle, better-auth, Tailwind v4+shadcn/ui. Use whenever the user mentions TanStack Start, asks to scaffold a full-stack Cloudflare app with SSR, wants an SSR dashboard, or asks for a React 19 + Cloudflare Workers app with file-based routing and server functions — even if they don't name TanStack Start specifically. No template repo — Claude generates every file fresh per ...

9881 votes

Pentest

PTES-aligned adversarial security audit for backend, frontend, and mobile applications. Produces a CVSS-scored Hacker Report with verified PoCs and phased remediation.

5491 votes
View all in development →