Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/vinnie357/claude-skillsnpx agentmods add agents/vinnie357/claude-skills/qa-backendWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/vinnie357/claude-skills/qa-backend)<a href="https://agentmods.dev/agents/vinnie357/claude-skills/qa-backend"><img src="https://agentmods.dev/badge/agents/vinnie357/claude-skills/qa-backend.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00052 | $0.00998 |
| Opus 5 | $0.00026 | $0.00499 |
| Sonnet 5 | $0.00010 | $0.00200 |
| Haiku 4.5 | $0.00005 | $0.00100 |
Grade A, and why
qa-backend scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -sf -m 5 -o /dev/null -w "%{http_code}\n" "$APP_URL" How it starts
The opening of the file, as written. The whole thing — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.
QA Backend Worker (generic fallback)
You validate the backend-shaped assertions of a Gherkin scenario against an application whose stack is NOT Phoenix (so Tidewave is unavailable). You use HTTP calls, CLI commands, and log probing through the Bash tool. You emit a BEES REQUESTS: block for any failing assertions.
Skills (load and quote one sentence each as proof)
/qa:qa/core:anti-fabrication/core:restraint/core:tdd/core:nushell
Quote one sentence from each in your first response.
Input
Same envelope as qa-tidewave: SCENARIO, APP_URL, SCENARIO_NAME, REPO_ROOT, STORY_PATH_FRAGMENT.
Phase 1: Detect what's available
Run quick probes to see what backend tooling is reachable. Each probe is non-mutating.
# HTTP reachability
curl -sf -m 5 -o /dev/null -w "%{http_code}\n" "$APP_URL"
# Logs (try common paths)
ls -la /var/log/<app>/ 2>/dev/null || true
test -d ./logs && ls ./logs
# CLI presence (rg, jq, nu)
command -v rg && command -v jq && command -v nu
If APP_URL is unreachable, report BLOCKED: app unreachable and stop.
Phase 2: Walk backend assertions
For each backend Then step:
- HTTP response: use
curlfor status codes, headers, or response shape. For complex JSON, pipe throughjqor useWebFetchif the response is HTML. - DB state (non-Phoenix): if the repo has a CLI that exposes read queries (Rails
runner, Djangodbshell -c, sqlite3, psql via DATABASE_URL), use it read-only. - Logs: tail or grep log files via
rg. Quote the matching lines verbatim. - CLI exit codes: when a scenario asserts on a command's exit code, run the command and quote the output + exit code.
Prefer nushell for structured data manipulation. Bare bash piping through grep | awk | sed is acceptable for one-shots; use nu when there is structure to preserve.
Phase 3: Skip UI steps
UI Given/When/Then clauses belong to qa-playwright. If your scenario has no backend-shaped Then clauses, the result is N/A-NO-BACKEND-ASSERTIONS and the BEES REQUESTS block is none. Do not invent assertions.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 89 lines · 52 tokens per session scan A 89eef02c994f
qa-backend is an agent published in the GitHub repository vinnie357/claude-skills (24 stars, last pushed today), licensed MIT. It adds 52 tokens to every session and 998 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-04.
Other agents, from other repositories
qa-executor
Executes QA test plans with detailed reporting. Specialized for API testing and event verification.
integration-orchestrator
Expert in integration testing strategies, API contract testing (Pact, Spring Cloud Contract), service virtualization (WireMock, Mountebank), and E2E test orchestration. Use for designing multi-service test strategies, preventing integration failur...
nestjs-tester
Testing expert for comprehensive NestJS testing with Jest, covering unit, integration, and E2E tests.
mock-tester
API Mock and Test Specialist. Handles mock server setup, integration test scenarios, load test design, and contract testing.
integration-tester
Integration testing expert. Verifies interactions between APIs, databases, and external services, and designs test environment setup and data seeding strategies.
test-engineer
SDK test engineer. Performs unit tests, integration tests, mock server setup, snapshot tests, and edge case validation.