Buyer eval · no sales call

One command. PASS or BLOCK.

sourcea-boot runs four checks before your agents execute — policy, provider, receipt freshness, queue truth. You get BOOT_REPORT.json on disk. Show the work, not a slide deck.

Clone sourcea-boot→ See live forensic proof

Proof checks — Live sample · checks passing

Boot gate — PASS or BLOCK · four checks

sourcea-boot — ★ GitHub eval

Status Live status Proof sample PDF

replay · sourcea-boot

Last factory boot · 2026-06-27 18:07:35 UTC · PASS

$ sourcea-boot --json
SOURCEA_BOOT PASS ok=true
  [PASS] policy_version: no policy file (POLICY.md) — skipped
  [PASS] provider: provider env present (ANTHROPIC_API_KEY)
  [PASS] receipt_fresh: no prior receipt — first boot allowed
  [PASS] queue_truth: no queue files configured — skipped
REPORT=/Users/sinakazemnezhad/Desktop/SourceA/BOOT_REPORT.json▋

Terminal injects from live boot-proof.json on deploy — same output you can verify on this page, no call required.

See live receipt

Run in 5 minutes

README path — clone · install · run.

01

Clone

git clone …/sourcea-boot

Public eval repo on GitHub
MIT license
Clone + editable install today
PyPI Phase 0b — not published yet

02

Run

sourcea-boot --json

Four checks · PASS or BLOCK
Writes BOOT_REPORT.json
Exit 0 = PASS · 1 = BLOCK

03

Prove

Self-serve proof

Open GitHub repo → Free proof sandbox

What the four checks mean

policy_version

Project policy / SSOT not stale vs last brief.

provider

LLM provider configured — no fake-green when keys exist.

receipt_fresh

Last boot/gate receipt fresh (<8h) and ok.

queue_truth

Agent queue head matches inbox truth when queue files present.