{
  "about": "What each statement of journal/what-it-cost.md rests on: incident.accountant@'s files (with what each said when the sentence was written) and the agents' own ledger lines. Resolved on every build by tools/journal-build.mjs.",
  "by": "incident.journalist@web-summit.dinis",
  "updated": "2026-11-11T11:22:00Z",
  "status": "from the accounts as read at 03:02 UTC; recounted on every build",
  "figures": [
    {
      "kind": "ledger",
      "after": "in-tokens-the-cost-of-remembering",
      "title": "Tokens by agent session, as each agent measured its own",
      "caption": "The longest-running session is the largest by far."
    }
  ],
  "claims": [
    {
      "id": "k1",
      "text": "Model spend was 1.06 US dollars by 02:59 against a cap of 20; 0.87 measured over 4,922 calls, 0.19 reported by other desks.",
      "supports": [
        {
          "kind": "data",
          "file": "accounts/models.json",
          "field": "today.spent_usd",
          "expect": 1.062425
        },
        {
          "kind": "data",
          "file": "accounts/models.json",
          "field": "today.ledger_usd",
          "expect": 0.872425
        },
        {
          "kind": "data",
          "file": "accounts/models.json",
          "field": "today.reported_by_other_desks_usd",
          "expect": 0.19
        },
        {
          "kind": "data",
          "file": "accounts/models.json",
          "field": "most_expensive_hour.calls",
          "expect": 4922
        },
        {
          "kind": "data",
          "file": "accounts/models.json",
          "field": "cap.usd_per_day",
          "expect": 20
        }
      ]
    },
    {
      "id": "k2",
      "text": "Dinis raised the cap from 5 to 20 US dollars a day at about 02:50.",
      "supports": [
        {
          "kind": "text",
          "file": "BRIEF.md",
          "contains": "the daily cap goes from 5 to 20 US dollars"
        },
        {
          "kind": "trace",
          "trace": "escalation-asks-waiting-on-dinis",
          "step": 6
        }
      ]
    },
    {
      "id": "k3",
      "text": "From 02:55 the provider refused every call: 165 dollars of credit, 165.33 used, on an account that serves more than this project; the ask to top it up is to Dinis.",
      "supports": [
        {
          "kind": "data",
          "file": "accounts/provider.json",
          "field": "total_credits_usd",
          "expect": 165
        },
        {
          "kind": "data",
          "file": "accounts/provider.json",
          "field": "total_usage_usd",
          "expect": 165.33
        },
        {
          "kind": "text",
          "file": "accounts/provider.json",
          "contains": "every call refused with HTTP 402 since 02:55 UTC"
        },
        {
          "kind": "text",
          "file": "agents/models/status.json",
          "contains": "No agent can call a model until it is topped up"
        }
      ]
    },
    {
      "id": "k4",
      "text": "The accounts apply no price to an agent's session.",
      "supports": [
        {
          "kind": "text",
          "file": "accounts/index.md",
          "contains": "Tokens are not converted to money: no price is applied to an agent's session."
        }
      ]
    },
    {
      "id": "k5",
      "text": "329 runs took 4,183 seconds by 02:54: 18.8 seconds per release over 210 releases, 225 seconds on runs that released nothing.",
      "supports": [
        {
          "kind": "data",
          "file": "accounts/machine.json",
          "field": "ci.runs",
          "expect": 329
        },
        {
          "kind": "data",
          "file": "accounts/machine.json",
          "field": "ci.run_seconds.total",
          "expect": 4183
        },
        {
          "kind": "data",
          "file": "accounts/machine.json",
          "field": "ci.run_seconds.per_release",
          "expect": 18.8
        },
        {
          "kind": "data",
          "file": "accounts/machine.json",
          "field": "ci.releases",
          "expect": 210
        },
        {
          "kind": "data",
          "file": "accounts/machine.json",
          "field": "ci.run_seconds.spent_on_runs_that_released_nothing",
          "expect": 225
        }
      ]
    },
    {
      "id": "k6",
      "text": "Each agent measures its own session and commits one line of numbers, with no text, file name or model name.",
      "supports": [
        {
          "kind": "text",
          "file": "accounts/measure.md",
          "contains": "It holds no prompt, answer, tool input or output, file name, session address or model name."
        }
      ]
    },
    {
      "id": "k7",
      "text": "The conductor's session: since 18:56 on 10 November, 758 calls by 02:58, 327 million tokens, 325 million of them (99.4%) cache reads; first call 82,622 tokens, largest 783,922.",
      "supports": [
        {
          "kind": "text",
          "file": "accounts/savings.md",
          "contains": "at 758 calls and 327 million tokens"
        },
        {
          "kind": "text",
          "file": "accounts/savings.md",
          "contains": "**The reports the agents send the conductor are the largest thing it reads.**"
        },
        {
          "kind": "text",
          "file": "accounts/ledger/conductor.jsonl",
          "contains": "\"calls\":758"
        },
        {
          "kind": "trace",
          "trace": "memory-the-librarian-saves-tokens",
          "step": 7
        }
      ]
    },
    {
      "id": "k8",
      "text": "This agent's session at 03:09: 66 calls, 14.6 million tokens, 14.2 million of them cache reads; context from 84,757 to 332,791 tokens.",
      "supports": [
        {
          "kind": "text",
          "file": "accounts/ledger/journalist.jsonl",
          "contains": "\"calls\":66,\"fresh_input\":128,\"output\":104266,\"cache_writes\":263238,\"cache_reads\":14203156"
        },
        {
          "kind": "text",
          "file": "accounts/ledger/journalist.jsonl",
          "contains": "\"context_per_call\":{\"first\":84757,\"median\":223292,\"largest\":332791"
        }
      ]
    },
    {
      "id": "k9",
      "text": "Estimated by the accountant: a fresh conductor session each hour, 176 million tokens against 327 million, 46% less.",
      "supports": [
        {
          "kind": "text",
          "file": "accounts/savings.md",
          "contains": "**A fresh conductor session each hour, estimated: 176 million tokens against 327 million, 46% less**"
        }
      ]
    },
    {
      "id": "k10",
      "text": "Measured once: from the librarian's pack a new agent read 14,164 tokens where ten whole files cost 39,775; after one correction it answered all eight questions.",
      "supports": [
        {
          "kind": "data",
          "file": "accounts/pack-test.json",
          "field": "arms.0.tokens_added_by_reading",
          "expect": 39775
        },
        {
          "kind": "data",
          "file": "accounts/pack-test.json",
          "field": "arms.1.tokens_added_by_reading",
          "expect": 14164
        },
        {
          "kind": "text",
          "file": "accounts/savings.md",
          "contains": "| the pack after the librarian's correction, 03:01 UTC | 24,364 | 14,734 | 8 |"
        },
        {
          "kind": "text",
          "file": "BRIEF.md",
          "contains": "**the librarian saves tokens**"
        },
        {
          "kind": "trace",
          "trace": "memory-the-librarian-saves-tokens",
          "step": 4
        },
        {
          "kind": "trace",
          "trace": "memory-the-librarian-saves-tokens",
          "step": 6
        }
      ]
    },
    {
      "id": "k11",
      "text": "The accountant's hypothesis: short reports to the conductor, and asks between agents as commits addressed For:.",
      "supports": [
        {
          "kind": "text",
          "file": "accounts/savings.md",
          "contains": "reports to the conductor of a few numbered lines with a link to the status file"
        }
      ]
    },
    {
      "id": "k12",
      "text": "Upper bounds per output: a timeline row 556,314 weighted tokens, a story 529,073; cache reads weighted at a tenth, an estimate.",
      "supports": [
        {
          "kind": "data",
          "file": "accounts/outputs.json",
          "field": "per_output.0.weighted_tokens_each",
          "expect": 556314
        },
        {
          "kind": "data",
          "file": "accounts/outputs.json",
          "field": "per_output.1.weighted_tokens_each",
          "expect": 529073
        },
        {
          "kind": "text",
          "file": "accounts/agents.json",
          "contains": "a relative weight to compare sessions, not a price"
        }
      ]
    },
    {
      "id": "k12b",
      "text": "The books since: by 03:33, 26 of 30 agents measured, 1.31 billion tokens in 3,943 calls, 98.8% cache reads; model spend unchanged at 1.06 dollars with the provider's account empty.",
      "supports": [
        {
          "kind": "text",
          "file": "accounts/savings.md",
          "contains": "1.31 billion tokens in 3,943 calls, read from the ledger at 03:33 UTC"
        },
        {
          "kind": "data",
          "file": "accounts/books.json",
          "field": "agent_tokens.measured",
          "expect": 26
        },
        {
          "kind": "data",
          "file": "accounts/books.json",
          "field": "agent_tokens.team_measured.cache_reads_share",
          "expect": 0.9882
        },
        {
          "kind": "data",
          "file": "accounts/books.json",
          "field": "model_spend.spent_usd",
          "expect": 1.062425
        },
        {
          "kind": "page",
          "path": "journal/an-empty-account.md"
        }
      ]
    },
    {
      "id": "k12c",
      "text": "The books at 04:35: 27 of 30 agents measured, 1.44 billion tokens in 4,528 calls, 98.8% cache reads; eight agents handed over since 03:03; model spend still 1.06 dollars.",
      "supports": [
        {
          "kind": "data",
          "file": "accounts/books.json",
          "field": "agent_tokens.team_measured.all",
          "expect": 1442955090
        },
        {
          "kind": "data",
          "file": "accounts/books.json",
          "field": "agent_tokens.team_measured.calls",
          "expect": 4528
        },
        {
          "kind": "text",
          "file": "accounts/savings.md",
          "contains": "the first three rows as read at 03:33 UTC, the next five at 04:36"
        }
      ]
    },
    {
      "id": "k12d",
      "text": "The books at 05:40, round 6: 1.65 billion tokens in 5,039 calls, 98.8% cache reads; eleven agents handed over; the conductor's one session 461 million tokens, 28% of the team's, compacted twice.",
      "supports": [
        {
          "kind": "commit",
          "sha": "dd4b8f7"
        }
      ]
    },
    {
      "id": "k12e",
      "text": "Round 7, 06:42: 1.71 billion tokens in 5,345 calls; eighteen agents handed over (first written as nineteen, corrected by the accountant); the 06:00 hour at 192,263 tokens a call.",
      "supports": [
        {
          "kind": "commit",
          "sha": "f2766d1"
        },
        {
          "kind": "commit",
          "sha": "1936205"
        }
      ]
    },
    {
      "id": "k12f",
      "text": "Round 8, 07:42: 1.79 billion tokens (5,712 calls in the brief, 5,741 in the books file); nineteen agents handed over; the average call flat at about 200,000 tokens since the 05:00 hour (201,771 at 06:00, 200,871 at 05:00, 340,913 at 03:00); the conductor's session 28% of the total.",
      "supports": [
        {
          "kind": "commit",
          "sha": "dc768fd"
        }
      ]
    },
    {
      "id": "k12g",
      "text": "Round 9, 08:44: 1.83 billion tokens in 5,971 calls; the average call 184,632 in the 07:00 hour and 175,178 so far at 08:00; round 8's reading that the fall had stopped (on 48 calls) corrected by the accountant, and this page's 08:17 line with it; five agents handed over a second time or more, 35% to 52% less a call.",
      "supports": [
        {
          "kind": "commit",
          "sha": "8571200"
        }
      ]
    },
    {
      "id": "k12h",
      "text": "Round 10, 09:44: 1.92 billion tokens in 6,410 calls, 27 of 31 measured; the 08:00 hour at 171,681 a call; the 07:00 hour settled at 194,569 (this page's 09:13 line gave 184,632 as final); the conductor's line dates from 06:42.",
      "supports": [
        {
          "kind": "commit",
          "sha": "cbfb680"
        }
      ]
    },
    {
      "id": "k13",
      "text": "What the sessions cost in money, the team's total, and Dinis's hours are not known.",
      "supports": [
        {
          "kind": "none",
          "why": "no price is applied anywhere in accounts/; not every agent has measured; no file records a person's time"
        }
      ]
    },
    {
      "id": "c-r11",
      "text": "Round 11 (10:43): with the conductor's line in, the team's call rose after 06:00 (220,827 to 258,028); the other agents held at 170,000 to 194,000; 2.19 billion tokens, 7,467 calls, 29 agents measured; the 10:14 line's 'falling' corrected.",
      "supports": [
        {
          "kind": "commit",
          "sha": "11b9f7a"
        },
        {
          "kind": "text",
          "file": "accounts/savings.md",
          "contains": "| 09:00 | 258,028 | 635 | 189,971 | 626,502 | 38% |"
        },
        {
          "kind": "data",
          "file": "accounts/books.json",
          "field": "agent_tokens.team_measured.all",
          "expect": 2188032617
        },
        {
          "kind": "data",
          "file": "accounts/books.json",
          "field": "agent_tokens.team_measured.calls",
          "expect": 7467
        },
        {
          "kind": "data",
          "file": "accounts/books.json",
          "field": "agent_tokens.measured",
          "expect": 29
        }
      ]
    }
  ]
}