Record native backend admission decision

This commit is contained in:
2026-11-15 23:38:22 +00:00
parent 176f59b176
commit 571ad70a54
2 changed files with 38 additions and 28 deletions

View File

@@ -4,7 +4,9 @@
remediation not implemented. D1 was resolved on 2026-11-15 and the architecture
policy updated; the focused cleanup audit is complete and implementation remains
pending. D2 was resolved on 2026-11-15: callers own report destination isolation;
the operator-documentation update remains pending.
the operator-documentation update remains pending. D3 was resolved on the same
date: native admission semantics remain in place; local scheduling improvements
await implementation planning.
This document owns classification, recommended
scope, and pending decisions for the [catalog](catalog.md). The source audits
retain their evidence; their suggested remedies are not approved requirements.
@@ -38,8 +40,8 @@ when substantial restructuring is appropriate.
**Ready** means suitable for implementation planning within the stated scope,
not implemented or validated by this document. **Investigate** denotes technical
work the agent can undertake without a product decision. **D3–D7** identify
pending substantive choices; D1 and D2 are recorded below as resolved. **Closed**
work the agent can undertake without a product decision. **D4–D7** identify
pending substantive choices; D1–D3 are recorded below as resolved. **Closed**
denotes no remediation for the specified scope. Mixed entries deliberately
separate ready work from decisions and exclusions.
@@ -98,7 +100,7 @@ behavior from proposed extensions.
| [27. Keep the validator execution class used by the runner](catalog.md#27-keep-the-validator-execution-class-used-by-the-runner) | C, A | Ready | Runtime and ledger can disagree for supported injected validator classes. Record the effective execution class with decisions and project it into the ledger. Preserve existing output fields and fallback behavior; no new validator registration policy. |
| [28. Verify signal and exit handling in the production executable](catalog.md#28-verify-signal-and-exit-handling-in-the-production-executable) | T | Ready | The startup helper duplicates the production boundary, so coverage does not prove actual executable signal wiring. Exercise representative startup and cancellation through the real executable or shared production entry point; keep narrower tests for other CLI behavior. |
| [29. Fail the release bootstrap when cleanliness cannot be established](catalog.md#29-fail-the-release-bootstrap-when-cleanliness-cannot-be-established) | C, T | Ready | The shell probe establishes that a failed status command can appear clean. Check its exit result explicitly and test that boundary. ReleaseKit already owns the independent delivery validation; do not add another local release validator or publication gate. |
| [30. Bound pending proposal work before backend admission](catalog.md#30-bound-pending-proposal-work-before-backend-admission) | E, A, T; P | Investigate implementation; D3 for policy | Fan-out and retained results scale with chunks; finite native admission rejection is itself working as configured. Design bounded local preparation/dispatch and concurrent result draining without changing backend ownership, stable identities, cancellation joins, or configured rejection semantics. Measure representative memory/contention to size the implementation. Guaranteeing a turn despite capacity rejection or adding a new scheduling control requires D3. |
| [30. Bound pending proposal work before backend admission](catalog.md#30-bound-pending-proposal-work-before-backend-admission) | E, A, T | Investigate implementation under resolved D3 | Bound local preparation/dispatch and drain results while preserving PromptKit admission ownership, configured waiting/rejection, stable identities, module barriers, and cancellation joins. Do not add an outer capacity-retry loop, duplicate backend limiter, or eventual-admission promise. The current scheduling-dependent selection of rejected chunks is not a compatibility requirement. Measure representative memory/contention to guide the implementation; any future admission-policy gap belongs upstream. |
| [31. Reduce peak memory for large transcript input and output](catalog.md#31-reduce-peak-memory-for-large-transcript-input-and-output) | E, A; P | Ready in part; D4 | Remove generic JSON round trips and release whole-input representations after their last use. The D1 audit additionally identifies source-debug reconstruction: write the acquired bytes directly and retain parsed/normalized views separately, without reopening the input. Share typed output projection; preserve encoding-failure and stream-delivery semantics. An input-size limit remains D4, and FIFO exclusion is not selected. |
| [32. Retain only the run evidence required by enabled outputs](catalog.md#32-retain-only-the-run-evidence-required-by-enabled-outputs) | E, A | Ready | The production consumers establish that much completed evidence is unused without its corresponding output. Retain only required live repair state, compact counts, and selected report/debug detail. Derive not-started counts without a full eager grid where possible; keep complete matrices when enabled and preserve fatal-origin evidence. |
| [33. Limit debug retention costs within a run](catalog.md#33-limit-debug-retention-costs-within-a-run) | E, A; P | Ready in part; D5 | Reduce duplicate event representations and serialization; coalesce equivalent warnings while retaining failure counts and affected evidence. Those are efficiency improvements. Quotas, automatic deletion, stopping future capture, or replacing self-contained interactions with reference-based artifacts change retained evidence and need D5. |
@@ -222,33 +224,37 @@ a workflow the caller already controls. The audit demonstrates overwriting when
callers share destinations; it does not establish an application promise of
per-run isolation. Correct writes to checked destinations in #1 remain independent.
## Resolved D3: retain native backend admission semantics
**Accepted:** 2026-11-15. **Applies to:** #30, with scheduling interactions in #15.
**Status:** decision recorded; local scheduling/resource improvements await
implementation planning.
PromptKit retains ownership of backend admission, including configured waiting
and recoverable capacity rejection. Use native settings where operators want
more waiting capacity. Audita will not add an outer admission-retry loop, a
separate backend limiter, or a guarantee that every chunk eventually receives
admission.
Improve Audita's local scheduling and memory use as engineering work: bound
pending preparation/dispatch, drain completed results, and avoid unnecessary
whole-transcript allocation. Preserve module barriers, stable identities,
ordered application, cancellation/join behavior, and native admission semantics.
The scheduling-dependent choice of chunks rejected by the current fan-out is
not a compatibility requirement; a different admission outcome caused by better
local scheduling does not itself establish a new admission policy.
Choosing to wait until every chunk can receive a turn would be a separate
processing policy. If reconsidered, first assess native PromptKit configuration
and seek upstream support for any gap. Native rejection is not a PromptKit defect.
Representative memory, contention, and cancellation checks should guide the
local implementation without creating a second admission system.
## Open Decisions
The five decisions D3–D7 below remain pending. Accepting or rejecting an optional
The four decisions D4–D7 below remain pending. Accepting or rejecting an optional
extension must not delay the independent ready portions of its catalog entries.
### D3. Should local dispatch change what backend saturation means?
**Applies to:** #30, with scheduling interactions in #15.
**Recommendation:** retain PromptKit's configured waiting and recoverable
capacity-rejection behavior. Improve local resource use and dispatch as
engineering work, but do not add an outer capacity-retry loop, a separate
backend limiter, or a promise that every chunk eventually receives admission.
Use native settings where operators want more waiting capacity.
**Alternative:** explicitly select an application policy that holds pending
chunks until they can receive a turn, subject to cancellation and existing
operation budgets. That could avoid losing useful work under finite queues, but
requires defining admission timing, fairness across proposal/validation/repair,
and whether deadlines cover the new wait. Seek suitable PromptKit support before
implementing local admission policy.
**Why approval:** replacing configured rejection with guaranteed waiting changes
run outcomes and timing. A bounded internal work window alone is not inherently
a policy change; its implementation must demonstrate that it stays on the chosen
side of this boundary. Native rejection is not a PromptKit defect.
### D4. Should Audita impose an application-level transcript input budget?
**Applies to:** the optional restriction in #31.

View File

@@ -301,7 +301,11 @@ comparison helper. The delivery tool has a separate prepared-checkout check.
One goroutine per chunk makes finite-capacity rejection depend on scheduling and
can retain requests and results across the whole run. Bound dispatch and result
retention without changing the module barrier or backend admission ownership.
Verify actual contention and cancellation in tests.
**Resolved D3:** preserve PromptKit's configured waiting and recoverable rejection.
Do not add an outer capacity-retry loop, duplicate backend limiter, or eventual-
admission promise. Current scheduling-dependent rejection order is not a
compatibility requirement. Verify memory use, actual contention, and cancellation
while improving local scheduling; any future admission-policy gap belongs upstream.
[Concurrency §5](04-concurrency.md#5-proposal-fan-out-makes-admission-under-finite-capacity-depend-on-scheduling) ·
[Resources §3](05-resources.md#3-backend-admission-leaves-chunk-fan-out-and-result-retention-unbounded) ·
[Testing §6](08-testing.md#6-concurrency-fixtures-leave-some-schedules-unproved)