--- name: smoke-test-09da259a6e0e description: >- Traces and verifies that something works end-to-end in any environment. Builds a check plan from natural language input, confirms it, then runs each check reporting pass/fail. Use when validating deployments, pipelines, features, or migrations. argument-hint: disable-model-invocation: true context: fork allowed-tools: - AskUserQuestion - Bash - Read - Glob - Grep - WebFetch - WebSearch metadata: trigger: "Validating deployments, pipelines, features, or migrations." tags: "testing, verification, devops" --- # Smoke Test Trace a resource, event, or feature through the system and verify it works end-to-end. ## Preferences _Read `~/.claude/skills/smoke-test/preferences.md` using the Read tool. If not found, no preferences are set._ ## Project context _On startup, use Bash to detect: CLAUDE.md presence, project stack files (`ls package.json Cargo.toml pyproject.toml go.mod requirements.txt`), and current git branch. Skip any that fail._ ## Command routing Check `$ARGUMENTS`: - **`help`** → display help then stop - **`config`** → interactive setup then stop - **`reset`** → delete `~/.claude/skills/smoke-test/preferences.md`, confirm, stop - **anything else** → run smoke test ### Help ``` Smoke Test — Trace and verify something works end-to-end Usage: /smoke-test Build and run a trace plan /smoke-test config Set environment and verbosity defaults /smoke-test reset Clear preferences /smoke-test help This help Examples: /smoke-test uploaded a file to workspace 30 in prod, check the pipeline /smoke-test verify login flow on staging /smoke-test did my migration run correctly in production? /smoke-test check webhook fires after order creation in dev How it works: 1. Reads project context (CLAUDE.md, repo structure, APIs) 2. Builds a trace plan — ordered checks 3. Confirms the plan with you 4. Runs each check, reporting pass/fail 5. Summarizes with diagnosis Current preferences: (shown above under Preferences) ``` ### Config Use **`AskUserQuestion`**: **Q1** — "Default environment?" (Production, Staging/Dev, Local, Ask each time (default)) **Q2** — "Trace output verbosity?" (Concise, Detailed, Verbose on failure (default)) Save to `~/.claude/skills/smoke-test/preferences.md`. ## First-time detection If no preferences file exists, show: "First time using /smoke-test? Run `/smoke-test config` to set default environment and verbosity, or continue — I'll auto-detect." Then proceed normally. ## Steps ### 1. Understand the project Read project context silently: - CLAUDE.md for conventions, endpoints, commands - Repo structure (`ls` top-level) - Tech stack from config files - API base URLs from CLAUDE.md, .env, or config files - Integration test files for endpoint patterns - Available MCP tools in the session ### 2. Parse the request From `$ARGUMENTS`, extract: - **What to verify**: Resource, event, or feature to trace - **Environment**: Where to check (use preference default or detect from input) - **Identifiers**: IDs, names, values to trace - **Expected behavior**: What "working" means If critical info is missing, ask via `AskUserQuestion`. ### 3. Build the trace plan Build an ordered list of checks. Each check: - **Check name**: Short description - **How**: Command or API call - **Expected**: What passing looks like - **Depends on**: Previous checks that must pass first See `reference/check-patterns.md` for heuristics on what to check. ### 4. Confirm the plan ``` Smoke test plan for: [what we're verifying] Environment: [env] Checks: 1. [Check name] — [how, briefly] 2. [Check name] — [how, briefly] 3. [Check name] — [how, briefly] Estimated: [N] API calls ``` Use **`AskUserQuestion`**: "Run this trace plan?" (Run all, Select specific, Add a check, Skip) ### 5. Execute Run each check sequentially (respecting dependencies): ``` [PASS] Check name Key details: value [FAIL] Check name Expected: X Got: Y [SKIP] Check name Depends on failed check ``` On failure: show expected vs actual, suggest likely cause, ask to continue or stop. Output patterns by verbosity preference: - **Concise**: pass/fail + one-liner - **Detailed**: full API responses - **Verbose on failure**: concise for pass, detailed for fail ### 6. Summary ``` Smoke Test Results: [what was verified] Environment: [env] [PASS] X/N checks passed [FAIL] Y/N checks failed [SKIP] Z/N checks skipped Failed: - [Check]: [one-line reason] Diagnosis: [Root cause or pattern if identifiable] Suggested next steps: - [Action per failure] ``` ## Principles - **Read-only by default**: Never write, mutate, or delete. This is a tracing tool. - **Be resourceful**: Use curl, gh, MCP tools, database CLIs, log files. - **Be adaptive**: Every project is different. Use CLAUDE.md and repo structure. - **Ask when stuck**: Rather than guess, ask. But try hard first. - **Show your work**: Show commands run and explain what you're checking. --- Source attribution: mostafa-drz Original project: https://github.com/mostafa-drz/claude-skills/blob/e289cedeca6f9d9d2fdd26ab43abc7ab3141c2cb/code/smoke-test/SKILL.md License: MIT License information: https://opensource.org/license/mit Original upstream instructions; downloaded from Mega AI Hub. ## Upstream license notice MIT License Copyright (c) 2026 Mostafa Darehzereshki Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. # Mega AI Hub resource metadata Creator / publisher: mostafa-drz Original project: https://github.com/mostafa-drz/claude-skills/blob/e289cedeca6f9d9d2fdd26ab43abc7ab3141c2cb/code/smoke-test/SKILL.md Attribution basis: Source repository record Fields of work: Development & Engineering Roles: QA / Test Engineer, DevOps / SRE, Backend Developer Specialties: Testing, DevOps Tools: Bash, Askuserquestion Country applicability: Country not specified Classification basis: Source category, role and specialization labels Instruction SHA-256: 7d751bbb1029072116c17ab43b549c12f29435ad437e63bf1c709717c52f4ac3