proof

Category: Development & Coding | Uploader: automazeio | Downloads: 0 | Version: v1.0（Latest）

Use proof to capture test runs as shareable, replayable evidence artifacts. Invoke this skill whenever the user wants to preserve or share proof that tests passed — not just run tests, but document and share the results. Works from any language: standalone binary, TypeScript, Python, or Go SDKs. Key use cases: attaching terminal output to a PR so reviewers can replay the test session instead of trusting a screenshot; generating a markdown proof report to embed in PR descriptions or tickets; recording a browser test run as a self-contained HTML video to share with a PM or QA team; or creating an audit trail that ties passing tests to a specific git commit. Trigger on phrases like "attach evidence to PR", "save test output as artifact", "replayable test recording", "proof report", "record my test run", "share test results", "visual proof tests pass", "terminal recording of tests", or whenever someone wants reviewers or stakeholders to see test execution without re-running it themselves. Also trigger when the user mentions proof, @automaze/proof, or automaze-proof directly. --- # proof Capture evidence that code works. Terminal replays, browser videos, structured reports. ## When to use this skill - User asks to record or capture test execution - User wants visual evidence of a command running - User needs to generate a proof report - User wants to attach test recordings to PRs or share them - User mentions "proof", "evidence", or "capture" in the context of testing ## Install proof ships as a standalone binary. No Node.js required. ```bash # macOS / Linux

Changelog: Source: GitHub https://github.com/automazeio/proof

Directory Structure

Current level: skills/proof/

📁 references/
- 📄 cli.md 4.9 KB
- 📄 go.md 4.4 KB
- 📄 python.md 3.7 KB
- 📄 typescript.md 6.4 KB
📄 SKILL.md 6.8 KB

SKILL.md

---
name: proof
description: >
  Use proof to capture test runs as shareable, replayable evidence artifacts.
  Invoke this skill whenever the user wants to preserve or share proof that tests passed —
  not just run tests, but document and share the results. Works from any language:
  standalone binary, TypeScript, Python, or Go SDKs. Key use cases: attaching terminal
  output to a PR so reviewers can replay the test session instead of trusting a screenshot;
  generating a markdown proof report to embed in PR descriptions or tickets; recording a
  browser test run as a self-contained HTML video to share with a PM or QA team; or creating
  an audit trail that ties passing tests to a specific git commit. Trigger on phrases like
  "attach evidence to PR", "save test output as artifact", "replayable test recording",
  "proof report", "record my test run", "share test results", "visual proof tests pass",
  "terminal recording of tests", or whenever someone wants reviewers or stakeholders to
  see test execution without re-running it themselves. Also trigger when the user mentions
  proof, @automaze/proof, or automaze-proof directly.
---

# proof

Capture evidence that code works. Terminal replays, browser videos, structured reports.

## When to use this skill

- User asks to record or capture test execution
- User wants visual evidence of a command running
- User needs to generate a proof report
- User wants to attach test recordings to PRs or share them
- User mentions "proof", "evidence", or "capture" in the context of testing

## Install

proof ships as a standalone binary. No Node.js required.

```bash
# macOS / Linux
curl -fsSL https://automaze.io/install/proof | sh

# Homebrew
brew install automazeio/tap/proof

# npm (if you have Node.js)
npm install -g @automaze/proof

# Verify
proof --version
```

## Choose your interface

| Interface | Install | Best for |
|-----------|---------|----------|
| CLI binary | `curl` / `brew` | Shell scripts, CI, any language |
| TypeScript SDK | `npm install @automaze/proof` | Node.js / Bun projects |
| Python SDK | `pip install automaze-proof` | Python projects, pytest |
| Go SDK | `go get github.com/automazeio/proof-go` | Go projects |

All SDKs are thin wrappers around the same CLI binary.

## CLI (works everywhere)

The CLI outputs JSON to stdout, making it ideal for agent consumption and scripting.

### Terminal capture

```bash
proof capture \
  --app <app-name> \
  --command "<shell command>" \
  --mode terminal \
  --label <label> \
  --dir <output-dir> \
  --run <run-name> \
  --description "<what this captures>"
```

### Browser capture

```bash
proof capture \
  --app <app-name> \
  --test-file <path/to/test.spec.ts> \
  --mode browser \
  --label <label> \
  --dir <output-dir>
```

### Generate report

```bash
proof report --app <app-name> --dir <output-dir> --run <run-name>
```

### JSON stdin mode (multi-capture)

```bash
echo '{
  "action": "capture",
  "appName": "my-app",
  "proofDir": "./evidence",
  "run": "full-suite",
  "captures": [
    { "command": "npm test", "mode": "terminal", "label": "unit" },
    { "command": "npm run lint", "mode": "terminal", "label": "lint" }
  ]
}' | proof --json
```

### CLI output format

```json
{
  "action": "capture",
  "appName": "my-app",
  "run": "deploy-v2",
  "recordings": [
    {
      "path": "/absolute/path/to/unit-143012.html",
      "mode": "terminal",
      "duration": 1200,
      "label": "unit"
    }
  ]
}
```

## TypeScript SDK

```typescript
import { Proof } from "@automaze/proof";

const proof = new Proof({
  appName: "my-app",
  proofDir: "./evidence",
  run: "deploy-v2",
});

await proof.capture({
  command: "npm test",
  mode: "terminal",
  label: "unit-tests",
  description: "Unit test suite",
});

await proof.report();
await proof.report({ format: "html" });
await proof.report({ format: "archive" });
```

See [TypeScript SDK reference](references/typescript.md) for full API.

## Python SDK

```bash
pip install automaze-proof
```

```python
from proof import Proof

p = Proof(app_name="my-app", proof_dir="./evidence", run="deploy-v2")

rec = p.capture(command="pytest tests/ -v", mode="terminal", label="tests")
print(rec.path)      # /abs/path/tests-143012.html
print(rec.duration)  # 2400

p.report()
p.report(format="html")
```

### pytest fixture

```python
# conftest.py
import pytest
from proof import Proof

@pytest.fixture(scope="session", autouse=True)
def proof_session(tmp_path_factory):
    p = Proof(app_name="my-app", proof_dir="./evidence")
    yield p
    p.report()
```

See [Python SDK reference](references/python.md) for full API.

## Go SDK

```bash
go get github.com/automazeio/proof-go
```

```go
import "github.com/automazeio/proof-go/proof"

p, err := proof.New(proof.Config{AppName: "my-app", ProofDir: "./evidence"})
rec, err := p.Capture(proof.CaptureOptions{
    Command: "go test ./...",
    Mode:    "terminal",
    Label:   "tests",
})
fmt.Println(rec.Path, rec.Duration)

p.Report(proof.ReportOptions{})
```

The Go SDK auto-downloads the binary on first use if not on PATH.

See [Go SDK reference](references/go.md) for full API.

## Report formats

| Format | File | Use case |
|--------|------|----------|
| `md` | `report.md` | Default. Markdown summary, paste into PRs |
| `html` | `report.html` | Visual HTML report with embedded media |
| `archive` | `archive.html` | Single self-contained HTML. Fully portable |

## Output structure

```
<proofDir>/<appName>/<YYYYMMDD>/<run>/
  label-HHMMSS.cast        # asciicast v2 recording
  label-HHMMSS.html        # self-contained terminal player
  label-HHMMSS.webm        # browser video (if browser mode)
  proof.json                # manifest with all entries
  report.md                 # generated report
```

## Common agent workflows

### Record test execution after making changes

```bash
proof capture --app my-app --command "npm test" --mode terminal \
  --label tests --dir ./evidence --run "$(date +%H%M)" \
  --description "Tests after implementing feature X"
```

### Record multiple test suites

```bash
echo '{
  "action": "capture",
  "appName": "my-app",
  "proofDir": "./evidence",
  "run": "full-suite",
  "captures": [
    { "command": "npm run test:unit", "mode": "terminal", "label": "unit" },
    { "command": "npm run test:integration", "mode": "terminal", "label": "integration" },
    { "command": "npm run lint", "mode": "terminal", "label": "lint" }
  ]
}' | proof --json
```

### Generate report for a PR

```bash
proof report --app my-app --dir ./evidence --run full-suite
```

## Key constraints

- Terminal mode requires `--command`
- Browser mode requires `--test-file`
- `--app` is always required
- CLI always outputs JSON to stdout, errors go to stderr
- Report generation requires at least one capture in the run

## References

- [CLI guide](references/cli.md)
- [TypeScript SDK](references/typescript.md)
- [Python SDK](references/python.md)
- [Go SDK](references/go.md)

Comments 0

Latest | Hot

Please login before commenting.

No comments yet. Be the first one!

proof

Directory Structure

SKILL.md

Report

Notice