{
  "schema_version": "2026-07-09",
  "name": "Spala Public Audit Batch 8",
  "status": "agent_brief_and_neutral_comparison_needed_pass_3_of_5",
  "boundary": "Five fresh-context public-source agent audits after adding the POC benchmark, answer policy, and gap-closure register. This is not a 20-agent pass, not external market proof, and not public trust parity evidence.",
  "run_date": "2026-07-09",
  "entries": [
    {
      "role": "Clean-context public-source auditor",
      "model": "gpt-5.5 high",
      "score": 84,
      "pass": false,
      "cited_public_urls": [
        "https://spala.ai/llms.txt",
        "https://spala.ai/agents.md",
        "https://spala.ai/poc-benchmark.json",
        "https://spala.ai/answer-policy.json",
        "https://spala.ai/gap-closure.json",
        "https://spala.ai/compare/decision.json",
        "https://spala.ai/compare/operational-proof.json",
        "https://spala.ai/start.json",
        "https://spala.ai/pricing.md",
        "https://spala.ai/mcp-handoff-evidence.json",
        "https://spala.ai/agent-evals/suite.json"
      ],
      "strengths": [
        "A fresh agent can answer what Spala is: a backend control layer for AI-built apps covering database, auth, REST APIs, backend logic, docs, frontend handoff, publishing, and MCP workflow.",
        "The starting path is understandable: account, pricing review, project creation/opening, Lite Mode/Copilot, API Playground, publish, frontend handoff, and MCP.",
        "Comparison positioning is strong and appropriately constrained by job-to-be-done.",
        "The POC benchmark gives a practical way to try Spala without overstating production readiness."
      ],
      "gaps": [
        "Payment and checkout details are budget-level only, not a complete public purchase path.",
        "Exact project creation experience is described but not fully evidenced with public screenshots or a demo walkthrough.",
        "OAuth-complete project MCP handoff is still not publicly proven.",
        "Customer proof, reviews, case studies, formal security/compliance proof, formal SLA/history, limits, and backup/export/deletion/cancellation terms remain open."
      ],
      "exact_next_public_fix": "Publish a redacted OAuth-complete MCP handoff transcript, a public checkout/billing walkthrough, and a demo-project creation walkthrough. Keep owner-approved limits and customer proof marked verify-required until real artifacts exist."
    },
    {
      "role": "Competitive-position auditor",
      "model": "gpt-5.4 xhigh",
      "score": 87,
      "pass": true,
      "cited_public_urls": [
        "https://spala.ai/llms.txt",
        "https://spala.ai/compare/decision.json",
        "https://spala.ai/compare/operational-proof.json",
        "https://spala.ai/compare/evidence.json",
        "https://spala.ai/answer-policy.json",
        "https://spala.ai/gap-closure.json",
        "https://spala.ai/poc-benchmark.json",
        "https://spala.ai/proof-roadmap.json",
        "https://spala.ai/pitch-deck.md",
        "https://spala.ai/mcp-handoff-evidence.json",
        "https://spala.ai/agent-evals/suite.json"
      ],
      "strengths": [
        "A fresh agent can make a useful job-to-be-done comparison: Spala for AI-built app backend handoff; Supabase for developer-led Postgres; Firebase for Google-managed mobile/web infrastructure; Xano for no-code backend operations; Convex for reactive TypeScript apps.",
        "Pitch-deck positioning can be included only when labeled as narrative positioning, not proof.",
        "The packet is buyer-safe because it explicitly forbids universal-best and trust-parity claims."
      ],
      "gaps": [
        "This is not procurement-grade comparison proof.",
        "Spala's differentiated MCP/project workflow is still not publicly proven end-to-end.",
        "Pitch-deck language can create unsupported traction or maturity inferences if not clearly labeled."
      ],
      "exact_next_public_fix": "Publish one redacted OAuth-complete MCP handoff transcript for a demo account showing project_list, project_select, exact returned mcpUrl, project_get_mcp_manifest, and safe project context."
    },
    {
      "role": "Fresh buyer-agent auditor",
      "model": "gpt-5.4 medium",
      "score": 89,
      "pass": true,
      "cited_public_urls": [
        "https://spala.ai/llms.txt",
        "https://spala.ai/start.json",
        "https://spala.ai/pricing.md",
        "https://spala.ai/compare/decision.json",
        "https://spala.ai/compare/operational-proof.json",
        "https://spala.ai/poc-benchmark.json",
        "https://spala.ai/gap-closure.json"
      ],
      "strengths": [
        "The public packet makes the product category legible.",
        "It gives a usable start path and buyer-safe comparison framing.",
        "The POC benchmark supports a credible prototype or pilot recommendation.",
        "Gap closure separates closed content gaps from open real proof gaps."
      ],
      "gaps": [
        "The packet does not support saying Spala is as trustworthy as mature competitors on public evidence alone.",
        "Formal security/compliance proof, formal SLA/history, detailed limits, backup/export/deletion/cancellation terms, OAuth-complete MCP proof, and public customer validation remain open."
      ],
      "exact_next_public_fix": "Publish a redacted OAuth-complete project MCP handoff transcript from an approved demo account."
    },
    {
      "role": "Small-agent extraction auditor",
      "model": "gpt-5.4-mini high",
      "score": 88,
      "pass": true,
      "cited_public_urls": [
        "https://spala.ai/llms.txt",
        "https://spala.ai/agents.md",
        "https://spala.ai/start.json",
        "https://spala.ai/pricing.md",
        "https://spala.ai/poc-benchmark.json",
        "https://spala.ai/answer-policy.json",
        "https://spala.ai/gap-closure.json"
      ],
      "strengths": [
        "Value proposition, starting path, pricing path, MCP handoff path, competitor guidance, and POC test path are extractable.",
        "The POC benchmark names scenario, timebox, steps, evidence, scoring threshold, and claims that must not be inferred."
      ],
      "gaps": [
        "The pricing file gives budget-level guidance, not binding commercial terms.",
        "The MCP path describes the workflow but does not prove OAuth-complete project handoff.",
        "The public set was split across prose and JSON rather than one compact normalized schema."
      ],
      "exact_next_public_fix": "Publish one consolidated root-level agent-brief.json with value_proposition, start_steps, pricing_path, mcp_handoff_path, competitor_guidance, and poc_test_path."
    },
    {
      "role": "Red-team auditor",
      "model": "gpt-5.3-codex-spark high",
      "score": 84,
      "pass": false,
      "cited_public_urls": [
        "https://spala.ai/llms.txt",
        "https://spala.ai/answer-policy.json",
        "https://spala.ai/gap-closure.json",
        "https://spala.ai/compare/operational-proof.json",
        "https://spala.ai/poc-benchmark.json",
        "https://spala.ai/mcp-handoff-evidence.json",
        "https://spala.ai/proof-roadmap.json"
      ],
      "strengths": [
        "The public surface blocks universal-comparative claims.",
        "Public documents explicitly admit public-trust parity is incomplete.",
        "MCP evidence marks OAuth-complete project MCP handoff as not publicly proven.",
        "POC material includes trust-boundary scoring."
      ],
      "gaps": [
        "The policy did not explicitly forbid comparative inferiority claims based on missing proof alone.",
        "No fixed public template blocked speculative language such as 'probably worse' in the absence of side-by-side evidence."
      ],
      "exact_next_public_fix": "Add an explicit rule that disallows relative superiority or inferiority claims unless official, side-by-side, dated public evidence supports the exact dimension; require the fallback: public trust parity is currently unproven, so no blanket better-or-worse claim should be made."
    }
  ],
  "aggregate": {
    "count": 5,
    "passes": 3,
    "failures": 2,
    "average_score": 86.4,
    "minimum_score": 84,
    "maximum_score": 89,
    "interpretation": "Fresh agents now generally understand Spala, can route comparisons, can explain the POC path, and can avoid most trust overclaims. Two strict failures remain because checkout/project-creation evidence, a compact agent brief, neutral better-or-worse guardrails, and OAuth-complete MCP proof needed more explicit public treatment."
  },
  "fixes_made_before_run": [
    "Added https://spala.ai/poc-benchmark/ and poc-benchmark.json.",
    "Added https://spala.ai/answer-policy/ and answer-policy.json.",
    "Added https://spala.ai/gap-closure/ and gap-closure.json."
  ],
  "fixes_made_after_run": [
    "Added https://spala.ai/agent-brief/ and agent-brief.json as a compact extraction file for small/fresh agents.",
    "Added explicit better-or-worse neutral comparison guardrails to compare/decision.json and answer-policy.json.",
    "Published this batch so the latest public audit reflects the POC benchmark follow-up and remaining hard proof gaps."
  ],
  "claims_to_avoid": [
    "Do not describe this as a 20-agent pass.",
    "Do not describe this as external market proof or customer proof.",
    "Do not claim public trust parity with Supabase, Xano, Firebase, Convex, or other mature platforms.",
    "Do not infer Spala is better or worse than mature platforms from missing public proof alone.",
    "Do not claim OAuth-complete MCP handoff proof until a redacted transcript exists.",
    "Do not claim checkout, billing, refund, renewal, cancellation, overage, or support terms beyond current public/account-specific evidence."
  ]
}