Close IR-WP-0004 and separate credential billing and research follow-ups

Assistant: codex
Assistant-Model: gpt-6-astra
Assistant-Session: 01a09cbc-910e-77a0-b336-3130194350b3
This commit is contained in:
tegwick 2026-09-14 15:56:57 +02:00
parent 955b9e871e
commit a72bc873b5
7 changed files with 199 additions and 40 deletions

View file

@ -1,40 +1,24 @@
# Custodian Brief — intelligence-radar
Updated: 2026-09-13. Domain infotech; topic custodian
(`cee7bedf-2b48-46ef-8601-006474f2ad7a`); prefix IR-WP-.
Repo ID: `3130bd49-a43f-4682-a25d-68b4d4de9b1f`.
Statehub: http://127.0.0.1:8000, primary railiance01.
Updated: 2026-09-14. Domain infotech; topic custodian.
## Current state
IR-WP-0001–0004 are finished. IR-WP-0004 closed at user request after transferring
unfinished scope into live follow-ups; its original incomplete tasks are cancelled
as transferred, not reported as successful.
v0.2 runs source capture, task-specific advice, outcome recording and frozen replay.
Daily/weekly jobs are configured; the bnt-lap001 user timer is enabled and its first
service invocation succeeded. Four primary pages and a dated two-profile price
comparison are retained. Real benchmark/access gaps are explicit; no production
success calibration is claimed. The complete ranking demo is synthetic and isolated.
- IR-WP-0005: credential delivery, coordinating SECRETS-WP-0010-T03. Existing Bao
key and three completed human approvals remain the starting point; native login/
delivery evidence is still pending. Do not repeat approvals or expose secrets.
- IR-WP-0006: billing reconciliation and durable settlement procedure. USD 0.023712
remains reserved in ir-openrouter-pilot-20260913; total authorization is USD 10.
- IR-WP-0007: comparable profile evidence, representative field outcomes,
publication and decomposition. Paid trials require both 0005 and 0006 readiness.
IR-WP-0001 through IR-WP-0003 are finished. IR-WP-0004 is active: broaden verified
profile evidence, collect real field outcomes, extend editorial/hosted operation.
Delivered: local v0.2 advisor/capture/scheduling, portable bundles, controlled trial
runner and no-network plans, public catalog adapter and retained source evidence.
40 tests and repository/archive checks pass. Public model listing does not prove
requester access or benchmark comparability; no real trial/calibration claimed.
Read INTENT → GOAL → SCOPE → AGENTS → workplans. docs/data-contracts.md describes the
CLI. make check validates; make demo-advisor demonstrates; make sync pushes committed
work and projects it to Statehub. See docs/evidence/v0.2-validation.md.
Resume checkpoint: docs/evidence/2026-09-13-resume.md. Portable bundles and the
USD 10 OpenRouter trial runner pass 32 tests. First inference and subsequent
key check returned HTTP 401; no completed real trials. USD 0.023712 remains
reserved pending reconciliation. Credential delivery blocks T02; T01/T03 continue.
2026-09-14: Existing Bao key confirmed by user. Access is blocked by the known
WARDEN-WP-0039-T03 caller/owner contract, not a request for a replacement key.
Native repair packet and dependencies: docs/evidence/2026-09-14-openrouter-access.md.
Latest: FLEX-WP-0026 supplies the native contract; SECRETS-WP-0010-T03 owns
remaining real approval/delivery. Radar `trials --plan` prepares exact pilot/full
review inputs without credentials or spending. 35 tests pass. See
docs/evidence/2026-09-14-trial-inputs.json and the adjacent trial plans.
Native delivery update: three human approvals are complete; do not repeat them.
Admin login/time remain blocked. Windows NT5DS cannot discover its domain time
peer; one-time resync did not fix it. Corporate domain connectivity and stable
host/guest time are prerequisites. See native-access-clock-check receipt.
Read INTENT → GOAL → SCOPE → AGENTS → docs/data-contracts.md → workplans.
Closure mapping: workplans/IR-WP-0004-evidence-coverage-and-field-calibration.md.
Evidence review: docs/evidence/2026-09-14-ir-wp-0004-review.md.

View file

@ -8,4 +8,5 @@ synthetic full-ranking/replay demonstration; tested outcome deduplication, stale
missing evidence handling, and durable scheduler retries; Statehub projection current.
Next: broaden verified benchmark/profile coverage and collect representative actual
outcomes. These research and calibration tasks continue in IR-WP-0004.
outcomes. These research and calibration tasks continue in IR-WP-0007; credential delivery
and billing are tracked independently in IR-WP-0005 and IR-WP-0006.

View file

@ -29,7 +29,7 @@ implemented; real execution is waiting on valid credentials. It does not learn
cross-task calibration, extract trustworthy
facts automatically, optimize decomposition, or estimate empirical cost/runtime
quantiles. Published costs are explicit workload scenarios. Catalog expansion and
actual outcome collection continue in IR-WP-0004.
actual outcome collection continue in IR-WP-0007.
Statehub holds work projections, not research evidence. Artifact hashes prove local
integrity, not truth. Scheduler operation depends on this workstation and its user

View file

@ -4,7 +4,7 @@ type: workplan
title: "Expand verified evidence and calibrate against field outcomes"
domain: infotech
repo: intelligence-radar
status: active
status: finished
flavor: implementation
owner: codex
topic_slug: custodian
@ -22,7 +22,7 @@ this work grows the evidence and operational reach without inventing calibration
```task
id: IR-WP-0004-T01
status: progress
status: cancel
priority: high
state_hub_task_id: "4f9538f6-3bf0-513f-baa0-5540430a2254"
```
@ -42,7 +42,7 @@ acceptance remains open.
```task
id: IR-WP-0004-T02
status: wait
status: cancel
priority: high
state_hub_task_id: "50522d33-d3e4-5a5e-ab8c-397a38c1ab17"
```
@ -72,7 +72,7 @@ The Bao key remains untested; no secret or policy changes were made.
```task
id: IR-WP-0004-T03
status: progress
status: cancel
priority: medium
state_hub_task_id: "e865449d-d53d-50bc-8983-2bd3605dfaee"
```
@ -122,3 +122,27 @@ The five vendor instruments do not match the original task selection or default
API harness. The 20 correlated controlled cases are a pilot, not representative
field calibration. T01/T03 remain progress; T02 remains wait on native delivery
and billing. No campaign/request changes, paid inference calls, or scope removals.
## Closure and residual handoff — 2026-09-14
User requested closure with credential delivery and billing separated. This is
closure of the delivered increment and transfer of unfinished scope, not a claim
that original research acceptance or native access has been achieved. Original
T01–T03 are `cancel` because their remaining execution moves to the live records
below; their implementation/evidence history above is preserved.
| Original scope | Authoritative follow-up |
| --- | --- |
| T01 comparable profiles and eligibility | IR-WP-0007-T01; credential evidence from IR-WP-0005 |
| T02 credential prerequisite | IR-WP-0005-T01/T02, coordinating SECRETS-WP-0010-T03 |
| T02 billing and pending reservation | IR-WP-0006-T01–T03 |
| T02 representative trials/calibration | IR-WP-0007-T02 |
| T03 publication/operations and decomposition | IR-WP-0007-T03/T04 |
Delivered: retained source/transcription evidence, public provider catalog adapter,
portable bundles, budgeted controlled runner and no-network trial plans, editorial
completion assessment, and 40 passing tests plus repository/archive integrity.
No completed real trial, verified provider spend or production calibration is
claimed. USD 0.023712 remains reserved; USD 10 remains the campaign authorization.
Credential work and billing work may progress independently; paid trials require
both readiness handoffs. Railiance Clock remains separate owner work.

View file

@ -0,0 +1,40 @@
---
id: IR-WP-0005
type: workplan
title: "Deliver admitted OpenRouter credentials to Radar"
domain: infotech
repo: intelligence-radar
status: ready
flavor: implementation
owner: codex
topic_slug: custodian
origin: residual
origin_ref: IR-WP-0004
created: "2026-09-14"
updated: "2026-09-14"
---
Created before closure of IR-WP-0004 at the user's request to separate credential
delivery from billing. Source files own the work; cross-repository owner plans
retain authority over native changes. Review baseline:
`docs/evidence/2026-09-14-ir-wp-0004-review.md`.
## Confirm native delivery prerequisites
```task
id: IR-WP-0005-T01
status: todo
priority: high
```
Coordinate with SECRETS-WP-0010-T03 and its existing owner records. Reuse the existing OpenBao key and completed human approvals; verify current native validity without repeating or extending approvals. Record exact recipient, caller/owner/tenant binding, admitted transport, login/time prerequisite and current owner evidence. Do not wait for billing reconciliation to perform credential-only diagnosis. No secret values in Git, logs or State Hub.
## Verify scoped delivery and return readiness
```task
id: IR-WP-0005-T02
status: wait
priority: high
```
Depends on T01 and native owner execution. Prove value-safe read-only provider key validation, correct recipient success, wrong caller/owner/tenant/path refusal, revocation and unchanged existing llm-connect delivery. Capture source/recipient revisions and sanitized native receipts. Do not widen policies or rotate the provider key merely to unblock Radar. This plan completes on admitted delivery evidence; paid inference and campaign accounting belong elsewhere.

View file

@ -0,0 +1,50 @@
---
id: IR-WP-0006
type: workplan
title: "Reconcile trial billing and establish campaign accounting"
domain: infotech
repo: intelligence-radar
status: ready
flavor: implementation
owner: codex
topic_slug: custodian
origin: residual
origin_ref: IR-WP-0004
created: "2026-09-14"
updated: "2026-09-14"
---
Created before closure of IR-WP-0004 at the user's request to separate credential
delivery from billing. Source files own the work; cross-repository owner plans
retain authority over native changes. Review baseline:
`docs/evidence/2026-09-14-ir-wp-0004-review.md`.
## Specify independent billing evidence and reconciliation contract
```task
id: IR-WP-0006-T01
status: todo
priority: high
```
Inventory the append-only events for ir-openrouter-pilot-20260913. Preserve the USD 0.023712 pending reservation and USD 10 total authorization. Define how provider billing evidence matches the failed request, what is unknown, who reviews a settlement, and how discrepancies are retained. HTTP 401 alone is not zero-charge proof. Accounting design can proceed independently of credential delivery; provider evidence retrieval may require an admitted read-only path.
## Implement and verify auditable reconciliation
```task
id: IR-WP-0006-T02
status: todo
priority: high
```
Provide a reviewed append-only settlement procedure or command. Preserve original reservations; reject duplicate/conflicting settlement, unknown reservation, invalid amount, cross-campaign references and budget overrun. Retain source billing evidence and reviewer attribution without credentials. Test uncertain/partial evidence and restart behavior. Do not fabricate a completed model trial to settle a failed request.
## Reconcile the outstanding reservation and hand off budget readiness
```task
id: IR-WP-0006-T03
status: wait
priority: high
```
Depends on T01/T02 and authoritative provider evidence. Resolve the USD 0.023712 reservation only to the supported actual amount, retaining any unresolved discrepancy explicitly. Rerender pilot/full plans for the same campaign and report spent, pending and remaining authorized budget. No campaign reset, additional authorization, or paid inference is implied. Completion requires reconciled evidence and a usable accounting handoff, not merely implementation tests.

View file

@ -0,0 +1,60 @@
---
id: IR-WP-0007
type: workplan
title: "Complete comparable evidence, field outcomes and publication"
domain: infotech
repo: intelligence-radar
status: ready
flavor: implementation
owner: codex
topic_slug: custodian
origin: residual
origin_ref: IR-WP-0004
created: "2026-09-14"
updated: "2026-09-14"
---
Created before closure of IR-WP-0004 at the user's request to separate credential
delivery from billing. Source files own the work; cross-repository owner plans
retain authority over native changes. Review baseline:
`docs/evidence/2026-09-14-ir-wp-0004-review.md`.
## Establish comparable execution-profile evidence
```task
id: IR-WP-0007-T01
status: todo
priority: high
```
Residual IR-WP-0004-T01. Pin a concrete task and five relevant benchmark versions, harnesses/tools, effort, attempts and source captures for at least two complete profiles. Resolve the vendor-instrument versus advisor-selection mismatch. Obtain requester access evidence from IR-WP-0005 before declaring eligibility. Produce the source-backed Frontier/Efficient comparison only when completeness and comparability checks pass. Preserve unknowns and conflicting claims; provider catalog metadata alone is insufficient.
## Collect and assess representative field outcomes
```task
id: IR-WP-0007-T02
status: wait
priority: high
```
Residual IR-WP-0004-T02. Paid execution depends on BOTH IR-WP-0005 credential readiness and IR-WP-0006 budget readiness; sampling-frame/verifier design can proceed earlier. Use the existing USD 10 campaign and reviewed pilot inputs. Define real task origins, inclusion/exclusion, held-out tasks, dependency groups, verifiers and failure taxonomy before claiming representativeness. The 20 close controlled variants are pilot evidence only. Retain successes/failures, actual costs/runtime and verifier artifacts; justify cross-task calibration and intervals before using them predictively.
## Complete reviewed publication and operational disposition
```task
id: IR-WP-0007-T03
status: todo
priority: high
```
Residual IR-WP-0004-T03. Build on portable bundles and the public catalog adapter. Define reviewed publication artifacts linking claims to captures, revisions and corrections. Keep generated drafts distinct from reviewed publication/notifications. Decide whether continuous uptime is required; retain the local timer unless a concrete requirement justifies a hosted worker and its live owner workplan.
## Compare decomposition with measured monolithic execution
```task
id: IR-WP-0007-T04
status: wait
priority: high
```
Depends on T02 matched held-out outcome evidence. Count orchestration, retries and verification in total cost/runtime, control task/model selection differences and retain uncertainty. No synthetic superiority or unsupported empirical quantiles. Carry any remaining research or infrastructure scope into live records before closure.