Report exact local outcomes from native execution purpose #105

What: Report exact local outcome usage from native execution purpose instead of whole-thread pooling.
Why: Purpose alone does not repair missing native response joins or prove complete usage.
So what: One exclusive consumer implementation starts after native source377 has a reviewed exact candidate.

Outcome

Supported task/trace/publish commands show a proved local Issue/PR outcome, native semantics, excluded shared usage and honest finality/price gaps. Canonical cost survives checkout cleanup. This advances Observer's explicit-ownership/native-evidence Objective.

Acceptance criteria

• Consume versioned immutable native purpose, delivery relations, causal edges and expected inventories, reusing existing assignments/reporting with provenance and minimal non-input witness.
• Literal Codex/Claude/OpenCode/Grok native input/response and child start/resume/background joins establish membership; display retouch, timestamp, current branch, reusable child id and thread links cannot. Preserve explicit non-T3 capture/Router semantics and surface conflicts.
• Native final usage and expected inventories reconcile before complete; removing a final response or child falsifies it despite terminal run status. Missing/native unsupported/pricing gaps remain partial or unknown, not zero.
• Issue-first/pre-PR and multiple-Issue planner scenarios stay separate with labeled PR delivery, not aliases or PR splits. Evaluate adjacent72 against its own contract before claiming it solved.
• Canonical reporting bypasses checkout. Legacy response-specific placement evidence persists at sync and survives cleanup without query-time Git; identical heads stay ambiguous.
• Exact local report exposes native identities and validated coverage/counter evidence for downstream export, without transcript content or manual assignment.

Blocked by

toolboxmd/chromeria#377 until its complete independently reviewed PR and exact native contract/proof handoff exists. Branch existence alone does not unblock.

Required proof

Meaningful native-id fixtures and isolated retained historical copies: two Issues, late PR, tool-free response, interruption, missing/late/mutable final records, native child resume/background, crash/import-order reversal, no query writes and removed checkout. Separate native-counter verification; real-task coverage across required harnesses, with original planner ledger absence explicit. Focused tests/current CI and independent exact-head review. Proven native inability returns a precise user scope decision, never partial-forever acceptance. Keep raw evidence private and avoid syncing this mixed parent into a misleading PR bill.

Parent

#58

Non-goals

No new transport, daemon, rates, manual ownership, invented historical purpose, subscription allocation, purpose repair or production deployment.

Elon record

Requirements and who asked: User approved the multiple-Issue accounting rule and requested building it; parent58 records the reviewed policy and native proof requirements.
Deleted: Whole-session substitutes, opaque registry, duplicate evaluator, manual capture and false full costs.
Bottleneck: Reviewed native membership/finality evidence at existing owners before downstream aggregation.
Checked myself: Live Observer main71bcb597a3ec8a15f3c595d043fc07d0e2271b59, native Chromeria3caa source, current report/fleet owners and adjacent72; required runtime proof is not yet passed.