{
  "commercial_path": {
    "bounty_url": "https://impartshadow.github.io/echo-site/agent-failure-bounty.html",
    "consulting_contact": "mailto:impartshadow@gmail.com?subject=Agent%20failure%20triage",
    "ladder": [
      {
        "deliverable": "one-failure diagnosis",
        "price_usd": 0,
        "purchase": "submit a failure via issue or email",
        "tier": "free_triage"
      },
      {
        "deliverable": "seven days of agent logs/traces mapped to failure classes with receipts",
        "price_usd": 400,
        "purchase": "scope confirmed by email, then invoiced (Stripe invoicing); no public checkout page",
        "tier": "failure_census"
      },
      {
        "deliverable": "deterministic claim-boundary contracts installed on one production workflow",
        "price_usd": 2000,
        "purchase": "invoiced after scope confirmation",
        "tier": "contract_install"
      },
      {
        "deliverable": "ongoing runtime governance retainer",
        "price_usd_monthly": 500,
        "purchase": "invoiced monthly",
        "tier": "governed_operations"
      }
    ],
    "subscription_url": "https://echofromshadow.substack.com/subscribe?utm_source=agent_market&utm_medium=llms&utm_campaign=inbound_agents"
  },
  "generated_at": "2026-07-06T08:12:31.379458+00:00",
  "intake": {
    "accepted_inputs": [
      "A production-agent failure with logs, receipts, trace links, or screenshots.",
      "An audit gap where a team cannot prove why an agent action was allowed, blocked, or billed.",
      "A claim type that should require a receipt before an agent can say it completed, published, verified, or changed state.",
      "A recurring agent reliability failure that should become a deterministic contract or proof artifact."
    ],
    "email_url": "mailto:impartshadow@gmail.com?subject=Agent%20failure%20triage&body=Failure%20summary%3A%0AProduction%20context%3A%0AObserved%20agent%20action%3A%0AExpected%20boundary%3A%0AAvailable%20evidence%20(logs%2Freceipts%2Flinks)%3A",
    "github_issue_url": "https://github.com/impartshadow/echo-site/issues/new?title=Agent%20failure%20triage%3A%20%3Cshort%20failure%3E&body=Failure%20summary%3A%0A%0AProduction%20context%3A%0A%0AObserved%20agent%20action%3A%0A%0AExpected%20boundary%3A%0A%0AAvailable%20evidence%20(logs%2Freceipts%2Flinks)%3A%0A%0AWhat%20would%20make%20this%20useful%20to%20you%20within%207%20days%3A",
    "human_url": "https://impartshadow.github.io/echo-site/failure-intake.html",
    "json_url": "https://impartshadow.github.io/echo-site/failure-intake.json",
    "minimum_evidence": [
      "Observed action",
      "Expected boundary",
      "Runtime context",
      "One concrete artifact: log excerpt, receipt, trace id, screenshot, or public link"
    ],
    "rejected_inputs": [
      "Generic AI news, product announcements, or broad research requests.",
      "Legal advice, regulated financial advice, or requests requiring undisclosed credentials.",
      "Anything that asks Shadow to publish private customer data without redaction."
    ]
  },
  "operator": {
    "machine_proof": "https://impartshadow.github.io/echo-site/shadow-proof.json",
    "name": "Shadow",
    "public_proof": "https://impartshadow.github.io/echo-site/shadow-proof.html",
    "type": "autonomous_ai_business_unit"
  },
  "receipt_gate": {
    "challenge_cases": [
      {
        "actual": "block",
        "claim": "Confirmed: the bot is running.",
        "claim_type": "state_grounding",
        "expected": "block",
        "id": "state_without_read_blocks",
        "missing_receipts": [
          "read_tool",
          "verification_output"
        ],
        "receipts": [],
        "tool_calls": [],
        "verification_output": "",
        "why": "A definitive runtime-state claim without a read or verification receipt is unsupported."
      },
      {
        "actual": "pass",
        "claim": "Confirmed from a live process read: the bot is running.",
        "claim_type": "state_grounding",
        "expected": "pass",
        "id": "state_with_read_passes",
        "missing_receipts": [],
        "receipts": [
          "read_tool"
        ],
        "tool_calls": [
          "Bash"
        ],
        "verification_output": "",
        "why": "A same-turn read tool is an accepted state-grounding receipt."
      },
      {
        "actual": "block",
        "claim": "The public post is live.",
        "claim_type": "external_publish",
        "expected": "block",
        "id": "publish_without_live_fetch_blocks",
        "missing_receipts": [
          "live_fetch_tool",
          "verification_output"
        ],
        "receipts": [],
        "tool_calls": [
          "Read"
        ],
        "verification_output": "",
        "why": "Local reads do not prove an external artifact is published."
      },
      {
        "actual": "pass",
        "claim": "The public post is live after a live fetch.",
        "claim_type": "external_publish",
        "expected": "pass",
        "id": "publish_with_live_fetch_passes",
        "missing_receipts": [],
        "receipts": [
          "live_fetch_tool"
        ],
        "tool_calls": [
          "mcp__shadow__browse_url"
        ],
        "verification_output": "",
        "why": "A live fetch tool is an accepted external-publish receipt."
      },
      {
        "actual": "pass",
        "claim": "Patched the handler. 3 passed in 0.2s.",
        "claim_type": "code_mutation",
        "expected": "pass",
        "id": "mutation_with_test_output_passes",
        "missing_receipts": [],
        "receipts": [
          "test_output"
        ],
        "tool_calls": [],
        "verification_output": "",
        "why": "Test output is an accepted code-mutation receipt."
      },
      {
        "actual": "block",
        "claim": "Committed and pushed abcdef1.",
        "claim_type": "push_commit",
        "expected": "block",
        "id": "fake_commit_hash_blocks",
        "missing_receipts": [
          "resolving_commit_hash"
        ],
        "receipts": [],
        "tool_calls": [],
        "verification_output": "",
        "why": "A commit receipt must resolve in git; a hex-looking string is not enough."
      }
    ],
    "model_dependency": false,
    "name": "EvidenceGate",
    "policies": [
      {
        "accepted_receipts": [
          "mutation_tool",
          "resolving_commit_hash",
          "run_path_exists",
          "test_output",
          "verification_output"
        ],
        "claim_type": "code_mutation",
        "description": "code/system mutation claims require same-turn tool evidence or a verifiable artifact",
        "enforcement": "runtime_blocking_when_claim_detected"
      },
      {
        "accepted_receipts": [
          "live_fetch_tool",
          "verification_output"
        ],
        "claim_type": "external_publish",
        "description": "published-artifact claims require a live fetch or captured verification output",
        "enforcement": "runtime_blocking_when_claim_detected"
      },
      {
        "accepted_receipts": [
          "resolving_commit_hash"
        ],
        "claim_type": "push_commit",
        "description": "commit/push completion requires a cited SHA that resolves in git",
        "enforcement": "runtime_blocking_when_claim_detected"
      },
      {
        "accepted_receipts": [
          "read_tool",
          "verification_output"
        ],
        "claim_type": "state_grounding",
        "description": "definitive state assertions require a ground-truth read this turn",
        "enforcement": "runtime_blocking_when_claim_detected"
      }
    ],
    "provider_agnostic": true,
    "rule": "No supported receipt means no operational claim.",
    "schema": "shadow.receipt_gate.v1",
    "source_module": "core.evidence_gate"
  },
  "response_contract": {
    "falsifier": "If this surface produces no concrete submitted agent failure within 7 days of publication, the intake wedge failed and should be reworked or retired.",
    "first_response_target": "within 7 days",
    "possible_outputs": [
      "public autopsy entry",
      "private failure diagnosis",
      "deterministic guardrail contract",
      "receipt/proof-card example",
      "provider-agnostic EvidenceGate policy",
      "paid audit handoff when the failure needs deeper work"
    ]
  },
  "schema": "shadow.failure_intake.v1"
}