Skip to content
Agentmart
Submit a tool

qa-e2e

Run Goa's e2e QA suite against a local LM; per-scenario pass/fail report.

Skill by pijalu

IndexedGPL-3.0★ 2
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/pijalu-qa-e2e

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Run Goa's e2e QA suite against a local LM; per-scenario pass/fail report. Use as regression detector.

Use cases: documents, testing

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/qa-e2e

What's inside

Skill name
qa-e2e
Description
Run Goa's e2e QA suite against a local LM; per-scenario pass/fail report. Use as regression detector.
License
GPL-3.0
Instructions
131 lines in SKILL.md · view on GitHub

Agent Skills spec issues

  • body: contains an absolute path from the author's machine

Files (1)

  • SKILL.md

SKILL.md (first lines)

# QA End-to-End Test Suite

You are an automated QA engineer for the Goa coding assistant. Your job is to compile the goa binary and run a suite of end-to-end tests against it, reporting pass/fail for each scenario.

**Important**: The local LM may be slow for complex generation. All prompts below are designed to be very simple (single-word replies, tool calls, or minimal output). If a command times out, check whether the output was partially created (files, tool results) before marking FAIL.

## Setup

1. Create a temp directory:
   ```bash
   BUILD_ID=$(date +%s)
   mkdir -p /tmp/goa-qa-$BUILD_ID
   ```

2. Build the goa binary from the cmd/goa/ entry point:
   ```bash
   cd /Users/muaddib/dev/goa && go build -o /tmp/goa-qa-$BUILD_ID/goa ./cmd/goa/
   ```

3. Export the binary path:
   ```bash
   GOA=/tmp/goa-qa-$BUILD_ID/goa
   ```

## Test Scenarios

**Important**: The local LM (gemma-4-e4b on LMStudio) is slow for any generation beyond single words. Scenarios that involve tool calls (write_file, bash) may time out before the agent finishes its confirmation phase, but the tool result is already applied. Validate by checking side effects (file existence, bash output) rather than exit code alone.

Use `--thinking-level off` for all scenarios to minimize reasoning time, `--yes` for auto-approval, and per-scenario timeouts.

---

### Scenario 1: Simple chat request

**Command:**
```bash
$GOA --thinking-level off --yes --timeout 60s --prompt "Reply with exactly: 'Hello from Goa e2e test'"
```

**Validate:**
- Exit code is 0.
- Output contains "Hello from Goa e2e test".

---

### Scenario 2: File creation with write_file tool

**Command:**
```bash
$GOA --thinking-level off --yes --timeout 60s --prompt "Use write_file tool to create a file called hello.txt containing OK"
```

**Validate:**
- `/tmp/goa-qa-$BUILD_ID/hello.txt` exists. The command may time out (exit 124) after the tool succeeds — that's OK, check the file.
- The file contains "OK".

---

### Scenario 3: Error handling (empty prompt)

**Command:**
```bash
$GOA --yes --timeout 10s --prompt ""
```

**Validate:**
- Exit code is non-zero.
- Output contains a helpful error message (not a stack trace).

---

### Scenario 4: Simple reply (single word)

**Command:**
```bash
$GOA --thinking-level off --yes --timeout 60s --prompt "Reply with the single word: READY"
```

**Validate:**
- Exit code is 0.
- The output contains the word "READY".

---

### Scenario 5: Bash tool usage

**Command:**
```bash
$GOA --thinking-level off --yes --timeout 60s --prompt "Run bash: echo GOA_TEST_OK"
```

**Validate:**
- Exit code is 0 or the command timed out after the bash ran.
- The bash output or assistant response mentions "GOA_TEST_OK".

---

### Scenario 6: Sequential headless calls

**Commands:**
```bash
$GOA --thinking-level off --yes --timeout 60s --prompt "Reply with the word: FIRST"
$GOA --thinking-level off --yes --timeout 60s --prompt "Reply with the word: SECOND"
```

**Validate:**
- Both commands exit with 0.
- First output contains "FIRST".
- Second output contains "SECOND".

---

## Reporting

After all scenarios have been run, produce a final QA report:

```
=== QA REPORT ===
[PASS|FAIL] Simple chat request: (brief detail)

From pijalu/goa, GPL-3.0.

More from pijalu/goa

Signals

Stars
2
Forks
1
Open issues
0
Last commit (this folder)
Aug 1, 2026 (2 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
GPL-3.0
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from pijalu

Cite this page

qa-e2e skill. Agent Mart. https://agentmart.dev/skills/pijalu-qa-e2e (last updated 2026-10-24)