{
  "schemaVersion": 2,
  "date": "2026.08.05",
  "publishedAt": "2026-08-05T21:36:28-07:00",
  "timeZone": "America/Los_Angeles",
  "title": "Assessing readiness for controlled internet observation",
  "publicationStatus": "Validated and published",
  "executiveSummary": [
    "Assessed whether Hiro is ready to begin the proposed controlled internet-observation probation and concluded that it is ready for implementation but not yet safe to start network observation.",
    "The main prerequisite exercise is complete: Hiro finished an end-to-end offline autonomous candidate cycle with frozen evidence, benchmark retention, strict rejection, no promotion, and Stage 6 still disabled.",
    "The live Hiro checkout is not yet a clean pinned baseline, and no dedicated internet-observation adapter, egress policy enforcement, sanitizer, immutable snapshot store, or snapshot-to-evaluation conversion pipeline is currently implemented.",
    "No internet access was enabled and no external request was made during this assessment."
  ],
  "workstreams": [
    {
      "title": "Prerequisite verification",
      "status": "Partially ready",
      "details": [
        "The first complete offline candidate cycle successfully exercised construction, target and regression tests, public and held-out evaluation, frozen evidence, benchmark-ledger retention, and the no-promotion boundary.",
        "The live Hiro HEAD remains c9148f1ec4cf9164e4256cc3ce93efdd6794282a and the Stage 6 DISABLED sentinel remains present.",
        "A clean isolated evaluation baseline exists, but it is not a complete replacement for the live checkout because ignored development-cycle modules were absent from that snapshot.",
        "The live checkout contains numerous existing modified and untracked files, so the autonomous clean-repository guard would block a routine probation launch from it."
      ]
    },
    {
      "title": "Internet-observation control plane",
      "status": "Not implemented",
      "details": [
        "Source inspection found design documentation for separating internet observation from evaluation and promotion, but no dedicated observation implementation.",
        "There is no current allowlisted GET/HEAD-only client that enforces domain, request-count, elapsed-time, response-size, content-type, and redirect budgets.",
        "There is no probation-specific private-network and local-address denial layer or redirect revalidation path.",
        "There is no implemented sanitizer and timestamped immutable snapshot packet format for treating retrieved content as untrusted evidence."
      ]
    },
    {
      "title": "Replay and evaluation integration",
      "status": "Not implemented",
      "details": [
        "No pipeline currently converts a real-world observation failure into a reviewable, deterministic evaluation case.",
        "No runner currently binds baseline and candidate evaluations to the same frozen web snapshot hash.",
        "The existing sandbox evaluator can provide the downstream construction, public and held-out evaluation, benchmark retention, and no-promotion boundary after snapshot-derived cases are safely introduced.",
        "The benchmark dashboard already has the autonomous-candidate review surface, but internet-observation provenance and snapshot details still need their own review fields."
      ]
    }
  ],
  "decisions": [
    "Do not begin live network observation until all specified controls exist and pass offline adversarial tests.",
    "Treat the current milestone as authorization to implement and validate the observation lane, not authorization to perform unrestricted browsing.",
    "Reconcile Hiro into an intentional clean pinned baseline before launching a timed probation.",
    "Keep internet discovery, candidate construction, evaluation, and promotion as separate authorities.",
    "Keep automatic promotion disabled and retain the Stage 6 DISABLED sentinel throughout implementation and probation."
  ],
  "validation": [
    {
      "check": "Live Git and Stage 6 state",
      "status": "passed",
      "result": "Confirmed live HEAD c9148f1ec4cf9164e4256cc3ce93efdd6794282a, extensive existing working-tree changes, and the Stage 6 DISABLED sentinel."
    },
    {
      "check": "Observation-lane implementation search",
      "status": "passed",
      "result": "Repository search found the previously published architecture but no dedicated network observation, sanitization, immutable snapshot, or snapshot-derived replay implementation."
    },
    {
      "check": "Hiro journal generation and frontend build",
      "status": "passed",
      "result": "The timestamped-entry test and production build completed successfully before publication."
    }
  ],
  "currentState": [
    "Hiro is ready to begin building the controlled internet-observation lane.",
    "Hiro is not yet ready to start the networked probation itself.",
    "No internet access is enabled, no external observation request has run, and no candidate has promotion authority.",
    "Stage 6 remains disabled."
  ],
  "limitations": [
    "The dirty live checkout prevents a clean routine launch and makes provenance harder to interpret until its changes are reconciled.",
    "The safety properties in the proposed probation are architectural requirements, not yet executable controls.",
    "Prompt-injection containment must be enforced structurally by separating untrusted retrieved content from instructions; a textual warning alone is insufficient.",
    "Snapshot-derived cases need review and deduplication rules so transient website failures do not automatically become misleading promotion evidence."
  ],
  "nextSteps": [
    "Reconcile the current Hiro changes into a reviewed, clean, pinned baseline.",
    "Implement the observation client, allowlist policy, private-network denial, redirect revalidation, hard budgets, sanitizer, immutable timestamped packets, and emergency stop.",
    "Add offline adversarial tests covering redirects, DNS and address changes, oversized or unsupported responses, prompt injection, malformed content, timeouts, and snapshot integrity.",
    "Implement reviewed conversion from frozen observations to deterministic cases and prove that baseline and candidate consume the identical snapshot hash.",
    "Only then launch a short read-only shadow probation with no credentials, state-changing methods, promotion, deployment, or Stage 6 activation."
  ],
  "disclosureNote": "This public entry contains no credentials, private held-out cases, personal data, or actionable details about unresolved security weaknesses."
}