Straits Industriesproduction
Search models, nodes, traces…⌘K
SI

gen_8cx9tovl0o

succeededbypass

chat request through @company/confidential — the complete decision record: guardrails, routing, fallback, cache and settlement.

Principal
Platform Admin(user)
Project
Core Platform
Profile
@company/confidential
Requested
@company/confidential
Resolved model
anthropic/claude-sonnet
Session
Tokens in / out
17 / 343
Cached tokens
15
Cost
$0.0052
Latency
1.53 s
Attempts
1
Age
12h ago

Decision timeline

  1. admission+0 ms

    chat request by prin_admin under profile @company/confidential (requested "@company/confidential")

  2. PRE_ROUTE+0 ms
    allowsize-budgetno findings [0.00ms]
  3. PRE_ROUTE+0 ms
    allowprompt-injection-heuristicno findings [0.00ms]
  4. PRE_ROUTE+0 ms
    allowegress-classifierno findings [0.00ms]
  5. PRE_MODEL+0 ms
    allowpattern-dlpno findings [0.00ms]
  6. PRE_MODEL+0 ms
    allowpii-redactorno findings [0.00ms]
  7. PRE_MODEL+0 ms
    redactsecrets-scannerrule=aws-access-key-id matches=1: AWS access key id detected [0.00ms]
  8. PRE_MODEL+0 ms
    restrict_routesecrets-scannerconfidential data => requireZdr, minTrustTier=B, regions=[ap-southeast-1] [0.00ms]
  9. cache+0 ms

    cache bypassed (policy off or non-cacheable request)

  10. route+0 ms

    selected anthropic/claude-sonnet via anthropic/acct_anthropic_ent_apse1 on node_sg_dc_01 (A-tier, ap-southeast-1) — objective=balanced, 12 candidates scored, 11 in fallback chain, 3 hard-filter rejections [router router_1.0, policy rp_confidential/@company/confidential@v6]

    15 candidates considered12 survived hard filters3 rejectedobjective: balancedfallback chain: 11router_1.0 · rp_confidential/@company/confidential@v6
    Hard-filter rejections
    • ep_sonnet_anthropic_use1@node_cf_use1_01region us-east-1 not in policy allowedRegions [ap-southeast-1]
    • ep_sonnet_anthropic_use1@node_sg_dc_01region us-east-1 not in policy allowedRegions [ap-southeast-1]
    • ep_sonnet_anthropic_use1@node_my_kul_office_01region us-east-1 not in policy allowedRegions [ap-southeast-1]
    Score breakdown — top 5
    #CandidatequalitycacheSavingscontractValuecompletionProbabilitycostlatencyPenaltyshadowPriceretryRiskaffinityPenaltyguardrailCostTotal
    1anthropic/claude-sonnet via anthropic · node_sg_dc_01 selected+0.558+0.017+0.024+0.396-0.012-0.061-0.029-0.0030.000-0.0400.851
    2anthropic/claude-haiku via anthropic · node_sg_dc_01 +0.5040.000+0.024+0.396-0.004-0.021-0.029-0.0030.000-0.0400.827
    3anthropic/claude-sonnet via anthropic · node_my_kul_office_01 +0.5580.000+0.024+0.388-0.012-0.063-0.029-0.0070.000-0.0400.819
    4anthropic/claude-haiku via anthropic · node_my_kul_office_01 +0.5040.000+0.024+0.388-0.004-0.023-0.029-0.0070.000-0.0400.812
    5anthropic/claude-sonnet via anthropic · node_my_kul_office_02 +0.5580.000+0.024+0.372-0.012-0.064-0.029-0.0180.000-0.0400.791
  11. attempt #1+0 ms

    #1 anthropic · account acct_anthropic_ent_apse1 · node node_sg_dc_01 — succeeded in 1530ms, 15 prompt-cached tokens

    succeededanthropicacct_anthropic_ent_apse1pool enterprisenode_sg_dc_01Tier C1.53 s

    17 in / 343 out / 15 cached

  12. POST_MODEL+0 ms
    allowpattern-dlpno findings [0.00ms]
  13. POST_MODEL+0 ms
    allowtoxicity-liteno findings [0.00ms]
  14. PRE_RETURN+0 ms
    allowpii-redactorno findings [0.00ms]
  15. PRE_RETURN+0 ms
    allowschema-validatorno findings [0.00ms]
  16. settlement+1530 ms

    settled $0.005156 — 17 in / 343 out / 15 cached tokens (360 total)

    Ledger entries
    • reserve$0.0063budget:bud_proj_corereserve est max for @company/confidential
    • settle$0.0052budget:bud_proj_coresettle actual provider usage (360 tok, internal charge $0.005456)
    • settle−$0.0052account:acct_anthropic_ent_apse1credit provider account acct_anthropic_ent_apse1 (double-entry counterpart)
    • refund−$0.0011budget:bud_proj_corerefund unused reservation
    Cost layers (memo §15.3)
    Provider list
    $0.0052
    Contract-adjusted
    $0.0047
    Internal charge
    $0.0055
    Contract savings
    $0.0004

    contract ct_anthropic_committed (committed)

  17. complete+1530 ms

    status=succeeded, model=anthropic/claude-sonnet, 1530ms end-to-end

fabric_metadata

The route-explanation envelope an API caller receives with X-OpenRouter-Metadata: enabled or routing.explain, reconstructed from this generation record.

fabric_metadata.json
{
  "requested": "@company/confidential",
  "resolved_model": "anthropic/claude-sonnet",
  "strategy": "guardrail-constrained-balanced",
  "policy_version": "rp_confidential/@company/confidential@v6",
  "router_version": "router_1.0",
  "provider": "anthropic",
  "region": "ap-southeast-1",
  "account_pool": "enterprise",
  "node": {
    "ref": "node_sg_dc_01",
    "trust_tier": "C"
  },
  "cache": {
    "response": "BYPASS",
    "prompt_affinity": "HIT",
    "predicted_cached_tokens": 15
  },
  "guardrails": [
    {
      "stage": "PRE_ROUTE",
      "action": "allow",
      "profile": "gp_strict",
      "evaluations": 3
    },
    {
      "stage": "PRE_MODEL",
      "action": "restrict_route",
      "profile": "gp_strict",
      "evaluations": 4
    },
    {
      "stage": "POST_MODEL",
      "action": "allow",
      "profile": "gp_strict",
      "evaluations": 2
    },
    {
      "stage": "PRE_RETURN",
      "action": "allow",
      "profile": "gp_strict",
      "evaluations": 2
    }
  ],
  "attempts": [
    {
      "provider": "anthropic",
      "status": 200
    }
  ],
  "routing_reason": [
    "selected anthropic/claude-sonnet via anthropic (balanced objective, score 0.8509)",
    "committed capacity preferred (contract economics)",
    "3 candidate(s) rejected by hard filters",
    "guardrails restricted the feasible route set (ZDR/tier/region)"
  ]
}