Workflow 0 — the Talomnia launch specification
Every figure on this page is summed from the execution rows below it. The trace is the record itself — not a report about the work, but the work, accounted unit by unit, including the units that failed validation and the ones that were rework.
- accounted units of execution
- 7
- wall-clock time recorded
- not measured
- units that were rework of an earlier attempt
- 0
Evidence record
- What is verified
- 7 execution units: 7 passed, 0 rejected, 0 reworked
- Revision
- Workflow record dated 2026-08-19
- Validation status
- done Pre-commercial self-use execution; completion is not independent validation.
The case in two minutes
- created here
- 1
- reused
- 1
- modified
- 1
- What the client asked
- How the Talomnia launch specification travelled from operator briefing to an accepted specification: client conversation → requirements → contradictions and TBDs → the specification document → validation → accepted specification → input to orchestration. This is where the specification ends and the building begins: everything built FROM this specification lives on its own page, «Building talomnia.com with Talomnia», and the market research on its own. This page used to carry all of it, and therefore called itself finished while the work was still running; the customer's R2-06 finding (issue #44) returned that, and the split is executed.
- Who executed
- 7 execution entries · 5 roles · 1 agent session
- Capabilities
- created: 1 · reused: 1 · modified: 1
- Checks
- validated: 7 · rejected: 0 · reworks: 0
- Result
- A specification is work that ends: the v1.3.1 launch specification was accepted by the operator and handed to orchestration, and that is why this page's status is «done». The building from it is not finished, and its own page says so with its own status. The split is not cosmetics: while both works shared one page, that page had to declare a single state for two works in different states — and the state it declared was false.
- Acceptance status
- done
- Coverage and currency
- 7 execution entries · snapshot
0a21ce0e8604
The entry count and digest state which set the totals were computed from. The digest is taken over EVERY figure in this block rather than over selected columns, so it changes when membership changes or when any figure beside it changes. An attestation narrower than its apparent scope is worse than none, because it buys trust it has not earned.
Every figure above is derived from the execution ledger and the artifact matrix below. The conclusions text is written by hand and is pinned to the entry count: if the ledger moves past it, the projection refuses to load rather than publish a stale claim.
Failures and rework
Derived directly from this case's execution ledger: no failed or reworked unit is hidden. Zero is a statement too.
Failed: 0 · Reworked: 0 · 7 of accounted units
- passed
- 7
- failed
- 0
- rework of an earlier attempt
- 0
Case economics
measured — from ledger rows with actual tokens and provider rates
estimated — reconstructed after the fact, with the method named in each row
measured share of the total
- Summed execution windows
- not measured — the sum of every entry’s window across parallel lanes — legitimately more than calendar time
- Active execution
- not measured — pure agent work, waits excluded
- Calendar span
- not measured — from the first start to the last finish
- Human participation
- 2.0h / $100.00 — recorded human review time and cost
The hours on this page are not billable hours: pre-commercially no billing is defined. Each figure above carries its own meaning and is never added to the others.
Methodology: "measured" are ledger rows whose cost derives from actual tokens and provider rates at recording time; "estimated" are rows reconstructed after the fact, with the basis named in the row itself (visible in the table below). The two sums are never blended into one figure.
Unmeasured, honestly: part of the orchestrator supervision time and human review outside ledger rows is not in these totals; infrastructure compute is near zero and recorded as zero.
We do not compare this cost against a human team: no defensible comparison methodology exists yet, and an invented number is worse than an honest absence.
Stage timeline
-
2026-08-19
client-conversation
knowledge-curator — operator briefing and the client conversation that framed what the launch has to be · public repository · not measured — this step predates the Workflow Ledger
-
2026-08-19
requirements
architect — requirements extracted from the conversation into numbered, checkable sections · public repository · not measured — this step predates the Workflow Ledger
-
2026-08-19
contradictions-and-tbd
evidence-auditor — contradictions and open TBDs listed as found — resolved with the operator, not silently smoothed over · public repository · not measured — this step predates the Workflow Ledger
-
2026-08-19
specification
architect — the launch specification document v1.3.1 — scope, sections, gates, mandates · public repository · not measured — this step predates the Workflow Ledger
-
2026-08-19
validation
evidence-auditor — the specification checked back against the requirements and against the open TBD list · public repository · not measured — this step predates the Workflow Ledger
-
2026-08-19
accepted-specification
customer-acceptance — operator acceptance of v1.3.1 as the specification the epic executes against · public repository · not measured — this step predates the Workflow Ledger
-
2026-08-19
orchestration-input
orchestrator-observer — the accepted specification handed to orchestration as the epic's input — execution from here is recorded on the «Building talomnia.com with Talomnia» workflow · public repository · not measured — this step predates the Workflow Ledger
Open the complete technical evidence Below is the full evidentiary layer: the Knowledge Contract, the artifact matrix with versions and statuses, the per-step time and cost ledger, links to commits and pull requests. Nothing is removed — only moved beneath the summary.
Task
How the Talomnia launch specification travelled from operator briefing to an accepted specification: client conversation → requirements → contradictions and TBDs → the specification document → validation → accepted specification → input to orchestration. This is where the specification ends and the building begins: everything built FROM this specification lives on its own page, «Building talomnia.com with Talomnia», and the market research on its own. This page used to carry all of it, and therefore called itself finished while the work was still running; the customer's R2-06 finding (issue #44) returned that, and the split is executed.
Knowledge Contract of the case
This page is itself published through a resolved Knowledge Contract, not a brief: K_id sha256:05cdee3c…, selected primary artifacts tal-blueprint-evidence-bearing-verification, tal-role-evidence-auditor, tal-skill-sanitization-gate, tal-skill-success-criterion-measurement; governance — tal-constraint-sanitized-projection, tal-policy-honesty-presentation. Embedded constraints: the epic rebuild must be disclosed, not silenced; an internal working abbreviation for the knowledge graph is never published — the public name is Professional Knowledge Graph. Acceptance criteria: page figures equal talo-ledger report; the artifact matrix is complete (type/version/source/status/task); every Atlas entity is backed by a graph revision.
TALO-0041 addendum (R2-06): the scope of this page is reduced to the specification arc. The epic's tasks and their accounting moved to the building page — not deleted, but attributed where the work actually happened.
Roles and participating agents
- roles
- Chief Architectcustomer-acceptanceevidence-auditorknowledge-curatororchestrator-observer
- constraint
- policy-honesty-presentation
Artifact matrix
Point at a column name, or tap it, and its definition appears here.
Artifact type — The managed knowledge-graph type an artifact belongs to: role, skill, blueprint plus the governance types constraint, success criterion, policy and capability description. The type decides which contract the artifact lives under and how it may be reused.
Artifact version — The version of the knowledge artifact as it took part in this task. Artifacts are versioned in the private knowledge repository and a task is bound to a specific version rather than to «current» — otherwise the evidence would stop reproducing at the next edit.
Artifact source — The artifact's identifier and the link to its Capability Atlas entry where one is published. The source is the address at which the claim can be checked; where no public entry exists the identifier is shown alone rather than an invented link.
Lifecycle status — The stage of a knowledge artifact: draft, reviewed, validated, shared-internal, public-sanitized, deprecated. Only public-sanitized artifacts reach a public surface; the rest are accounted for as withheld rather than quietly omitted.
Disposition — How a knowledge artifact took part in the task: created — authored for the first time, reused — used unchanged, modified — revised. The reuse metric stands on this mark, so it is recorded in the ledger entry rather than inferred afterwards.
Linked task — The identifier of the task in which this artifact was touched. A task is the unit of project work; it contains the subtasks that time and cost are accounted against, and this identifier is what ties a row of evidence to the ledger records.
| — | — | Projects/Talomnia/docs/TZ-Talomnia-Launch-v1.3.1.md | — | created | TALO-0000 |
| — | — | Projects/Talomnia/docs/TZ-Talomnia-Launch-v1.3.1.md | — | reused | TALO-0000 |
| — | — | Projects/Talomnia/docs/TZ-Talomnia-Launch-v1.3.1.md | — | modified | TALO-0000 |
The matrix shows which knowledge artifacts participated in the task and how: created — made for the first time, reused — reused unchanged, modified — reworked.
Time and cost
Point at a column name, or tap it, and its definition appears here.
Subtask — The accounted unit of execution: one step of a task with its own Workflow Ledger record carrying time, cost and validation status. A row of the time-and-cost table is a subtask, not a whole task.
Role — A function and area of responsibility within a specific piece of work — for example Developer, Architect, Reviewer, Tester. A managed type of the knowledge graph: a role carries versions, relations, provenance and evidence.
Wall time — The time from the start of a unit of work to its finish, including waiting, pauses and everything in between. It is the elapsed duration of the execution window, not the amount of work done; across parallel lanes the sum of windows is legitimately larger than the calendar time.
Active time — The time during which work was actually executing, with the waiting and the idle stretches inside the window removed. It is always less than or equal to wall time; the difference between them is waiting, not lost work.
Model cost — The cost of the language-model calls for this unit of work: input and output tokens at the provider's price. Where no metered provider invoice exists for the session, the figure is marked as an estimate and its cost basis is published beside the row.
Compute cost — The cost of machine time outside the model: servers, builds, test runs and other infrastructure spent on this unit of work. A zero here means measured absence of such spend, not absence of measurement.
Human cost — The human time on this unit of work, converted to money at the rate the operator declared. Both the rate and the time itself are reconstructed after the fact from evidence rather than measured while the work ran, so every figure in this column is marked as reconstructed — and a row with no human participation has no human cost at all.
Total cost — The sum of model, compute and human cost for this unit of work. Only what was measured is summed: if nothing in the column was measured the total is not zero — it is unknown, and says so.
| TALO-0000/accepted-specification ~est | customer-acceptance | — | — | — | — | — | — |
| TALO-0000/client-conversation ~est | knowledge-curator | — | — | — | — | — | — |
| TALO-0000/contradictions-and-tbd ~est | evidence-auditor | — | — | — | — | — | — |
| TALO-0000/orchestration-input ~est | orchestrator-observer | — | — | — | — | — | — |
| TALO-0000/requirements ~est | architect | — | — | — | — | — | — |
| TALO-0000/specification ~est | architect | — | — | — | — | 87.50 | — |
| TALO-0000/validation ~est | evidence-auditor | — | — | — | — | 12.50 | — |
| Grand total | — | — | — | — | 100.00 | — |
Wall time runs from start to finish including all waiting; active time is pure execution. Costs are actual or estimates with a recorded basis (cost basis in the ledger).
Figures marked «reconstructed» were derived after the fact from the evidence named in the basis, not metered while the work ran. They are estimates, and the basis is published so a reader can disagree with them.
Human time is converted to money at a declared rate of $50/hour, set by the operator. That rate is an assumption, not a measured or billed price, and the time it multiplies is itself reconstructed — so every human figure here is a reconstruction of a reconstruction, and is marked as one. Rows with no human time carry no human cost: a task no person worked on is not a task a person worked on for free.
- reconstructed — R E C O N S T R U C T E D , n o t m e a s u r e d . T h i s s t e p p r e d a t e s t h e W o r k f l o w L e d g e r ( f i r s t e n t r i e s 2 0 2 6 - 0 8 - 1 9 ) , s o n o s e s s i o n t e l e m e t r y e x i s t s : w a l l t i m e , a c t i v e t i m e , t o k e n s a n d m o d e l c o s t a r e r e c o r d e d a s n u l l — n o t m e a s u r e d — r a t h e r t h a n a s 0 . O n l y t h e c a l e n d a r d a t e i s k n o w n , s o s t a r t e d _ a t a n d f i n i s h e d _ a t b o t h c a r r y t h e d a t e a n d t h e c a l e n d a r s p a n i s r e p o r t e d a s u n k n o w n . human_cost_usd is null (not measured): the founder's time on the founding specification was not billed and the site publishes no hourly rate for a human, so no defensible rate exists to convert this time into money. Writing 0 would say the work was free.
- reconstructed — R E C O N S T R U C T E D , n o t m e a s u r e d . T h i s s t e p p r e d a t e s t h e W o r k f l o w L e d g e r ( f i r s t e n t r i e s 2 0 2 6 - 0 8 - 1 9 ) , s o n o s e s s i o n t e l e m e t r y e x i s t s : w a l l t i m e , a c t i v e t i m e , t o k e n s a n d m o d e l c o s t a r e r e c o r d e d a s n u l l — n o t m e a s u r e d — r a t h e r t h a n a s 0 . O n l y t h e c a l e n d a r d a t e i s k n o w n , s o s t a r t e d _ a t a n d f i n i s h e d _ a t b o t h c a r r y t h e d a t e a n d t h e c a l e n d a r s p a n i s r e p o r t e d a s u n k n o w n . RECONSTRUCTED human time. Basis: the specification document (Projects/Talomnia/docs/TZ-Talomnia-Launch-v1.3.1.md, sha256 230bab820a37d554…) records its own revision history in
revision_note— v1.3 lists six named operator decisions on § 15 plus the § 15 mode change, and v1.3.1 records the resolution of one QA blocker: eight documented operator decisions in total. Assumption, stated so it can be disputed: 15 minutes of operator deliberation per documented decision. The COUNT is measured from the document; the 15 minutes is an assumption, not a measurement. Document volume was rejected as the basis: word count measures the artifact, not the human judgement — the same 8,000 words could carry one decision or fifty. human_cost_usd is null (not measured): the founder's time on the founding specification was not billed and the site publishes no hourly rate for a human, so no defensible rate exists to convert this time into money. Writing 0 would say the work was free. Human cost is DERIVED, not measured: recorded human time × a rate of $50/hour declared by the operator (issue #50 R6-04). The rate is an operator decision, not a market or billed price, and the time it multiplies is itself reconstructed — so this figure is a reconstruction resting on a reconstruction, and is published as one.
Links to logs and commits
- https://github.com/Arcanada-one/talomnia-trace/tree/main/workflows/talomnia-launch-specification
- https://github.com/Arcanada-one/talomnia-trace/blob/main/workflows/talomnia-launch-specification/artifacts/README.md
- https://github.com/Arcanada-one/talomnia-trace/blob/main/workflows/talomnia-launch-specification/correspondence/README.md
- https://github.com/Arcanada-one/talomnia-trace/blob/main/workflows/talomnia-launch-specification/tasks/README.md
Contradictions and open questions
Conclusions
A specification is work that ends: the v1.3.1 launch specification was accepted by the operator and handed to orchestration, and that is why this page's status is «done». The building from it is not finished, and its own page says so with its own status. The split is not cosmetics: while both works shared one page, that page had to declare a single state for two works in different states — and the state it declared was false.