Aveto runs
Every run, read from its own records.
One row per run in gpatwa/aveto-support. A script reads each run's state file, trace, end review and event log; nothing here is typed by hand, and nothing on this page can change a run. A figure a run's records do not state is shown as not recorded.
13runs recorded
13marked done
11,941ktokens, peak context summed per spawn (self-reported)
6ended over their plan total
| Run | Status | Spend | Cap | Tier | Charter | Event log | Pull request | Asked after Charter | Open for owner |
|---|---|---|---|---|---|---|---|---|---|
| result-page-clarity 2026-10-27 |
done pull request opened 2026-10-27; live on revision ca-aveto-ask--0000014; owner reviews and merges |
581k of 400k (145% of plan) self-reported |
800k hard cap | 3 | 8eacc4078037approved 2026-10-27 |
intact · 387 lines | not recorded | 0 | 3 |
| polished-chat-experience 2026-10-26 |
done pull request 11 merged 2026-10-27; live on revision ca-aveto-ask--0000011 |
1,570k of 1,000k (157% of plan) self-reported |
not recorded | 3 | a24c3c56e430approved 2026-10-26 |
intact · 1043 lines | #11 | 0 | 6 |
| fix-service-accounting 2026-10-23 |
done pull request 9 merged 2026-10-24 |
919k of 730k (126% of plan) self-reported |
not recorded | 3 | afc984d1approved 2026-10-23 |
intact · 533 lines | #9 | not recorded | not recorded |
| public-ask-the-docs 2026-10-20 |
done pull request #5 open; the merge is the owner's act |
2,496k of 2,000k (125% of plan) self-reported |
not recorded | 3 | e04b82539e1fapproved 2026-10-23 |
no event log | #5 | not recorded | 3 |
| land-answerability 2026-10-17 |
done PR opened; owner merges |
651k of 450k (145% of plan) self-reported |
not recorded | 3 | not recorded | no event log | #4 | not recorded | not recorded |
| docs-answerability-2 2026-10-12 |
done | 2,773k of 2,600k (107% of plan) self-reported |
3,400k ceiling | 3 | not recorded | no event log | not recorded | not recorded | not recorded |
| docs-abstention 2026-10-09 |
done | 672k of 780k (86% of plan) self-reported |
not recorded | 2 | not recorded | no event log | not recorded | not recorded | not recorded |
| docs-retrieval-4 2026-10-03 |
done | 581k of 585k (99% of plan) self-reported |
not recorded | 2 | not recorded | no event log | not recorded | not recorded | not recorded |
| docs-retrieval-3 2026-10-01 |
done | 314k of 360k (87% of plan) self-reported |
not recorded | 2 | not recorded | no event log | not recorded | not recorded | not recorded |
| docs-retrieval-3-proof 2026-10-01 |
done | 143k of 240k (60% of plan) self-reported |
not recorded | 2 | not recorded | no event log | not recorded | not recorded | not recorded |
| docs-retrieval-2 2026-09-29 |
done | 312k of 460k (68% of plan) self-reported |
not recorded | 2 | not recorded | no event log | not recorded | not recorded | not recorded |
| docs-retrieval-2-proof 2026-09-29 |
done | 75k of 430k (17% of plan) self-reported |
not recorded | 2 | not recorded | no event log | not recorded | not recorded | not recorded |
| docs-retrieval 2026-09-27 |
done | 854k of 1,130k (76% of plan) self-reported |
not recorded | 2 | not recorded | no event log | not recorded | not recorded | not recorded |
Where the tokens went, by role
Measured from the logs Claude Code itself writes, for 13 of 13 runs. Subagent stages only: the Orchestrator’s own turns are not included, so these totals understate a run. “Processed” counts every token a stage handled, cache reads included; “input-equivalent” weights them by price ratio. Neither is a bill. A run with no usage.json is left out, not estimated.
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| software-architect | 19 | 579 | 70.6M | 17.0M | 31% |
| qa-evidence | 17 | 507 | 37.3M | 7.2M | 17% |
| backend-architect | 14 | 391 | 37.1M | 7.7M | 16% |
| security-privacy | 9 | 268 | 25.2M | 5.0M | 11% |
| release-manager | 11 | 196 | 18.3M | 5.0M | 8% |
| ai-engineer | 5 | 117 | 9.1M | 2.0M | 4% |
| tech-writer | 15 | 129 | 7.3M | 2.5M | 3% |
| engineering-manager | 10 | 102 | 4.8M | 2.1M | 2% |
| frontend-developer | 4 | 70 | 4.7M | 1.3M | 2% |
| compliance-reviewer | 4 | 55 | 4.1M | 1.4M | 2% |
| post-launch-learning | 9 | 56 | 2.6M | 1.4M | 1% |
| ai-governance | 2 | 15 | 0.9M | 0.4M | 0% |
| cloud-deployment | 1 | 17 | 0.9M | 0.3M | 0% |
| production-verification | 4 | 24 | 0.8M | 0.5M | 0% |
| ui-designer | 3 | 16 | 0.8M | 0.5M | 0% |
| sre | 1 | 8 | 0.3M | 0.2M | 0% |
| product-manager | 2 | 8 | 0.1M | 0.1M | 0% |
| ux-researcher | 1 | 3 | 0.1M | 0.1M | 0% |
result-page-clarity 6.7M processed · 8 of 8 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| release-manager | 1 | 26 | 2.3M | 0.6M | 35% |
| software-architect | 1 | 23 | 1.3M | 0.4M | 19% |
| qa-evidence | 1 | 21 | 1.1M | 0.3M | 16% |
| frontend-developer | 1 | 19 | 1.0M | 0.3M | 15% |
| post-launch-learning | 1 | 7 | 0.3M | 0.3M | 5% |
| security-privacy | 1 | 7 | 0.3M | 0.1M | 4% |
| production-verification | 1 | 5 | 0.2M | 0.2M | 3% |
| tech-writer | 1 | 5 | 0.2M | 0.1M | 3% |
polished-chat-experience 35.7M processed · 16 of 16 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| software-architect | 1 | 67 | 13.1M | 2.6M | 37% |
| backend-architect | 4 | 79 | 9.1M | 1.7M | 25% |
| qa-evidence | 1 | 44 | 3.5M | 0.7M | 10% |
| security-privacy | 1 | 29 | 2.5M | 0.6M | 7% |
| frontend-developer | 2 | 31 | 2.5M | 0.7M | 7% |
| release-manager | 1 | 25 | 2.2M | 0.5M | 6% |
| tech-writer | 2 | 18 | 1.2M | 0.3M | 3% |
| ui-designer | 1 | 9 | 0.7M | 0.4M | 2% |
| engineering-manager | 1 | 9 | 0.5M | 0.2M | 1% |
| post-launch-learning | 1 | 5 | 0.2M | 0.1M | 1% |
| production-verification | 1 | 7 | 0.2M | 0.1M | 1% |
fix-service-accounting 15.2M processed · 10 of 10 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| software-architect | 1 | 47 | 7.2M | 1.4M | 47% |
| security-privacy | 1 | 29 | 2.6M | 0.6M | 17% |
| backend-architect | 1 | 28 | 2.0M | 0.8M | 13% |
| release-manager | 1 | 11 | 0.9M | 0.3M | 6% |
| qa-evidence | 1 | 20 | 0.9M | 0.3M | 6% |
| tech-writer | 2 | 14 | 0.7M | 0.3M | 5% |
| engineering-manager | 1 | 8 | 0.4M | 0.2M | 3% |
| production-verification | 1 | 7 | 0.3M | 0.1M | 2% |
| post-launch-learning | 1 | 4 | 0.2M | 0.1M | 1% |
public-ask-the-docs 47.0M processed · 31 of 31 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| software-architect | 5 | 112 | 13.7M | 3.2M | 29% |
| security-privacy | 1 | 55 | 8.3M | 1.6M | 18% |
| backend-architect | 4 | 84 | 5.7M | 1.6M | 12% |
| qa-evidence | 1 | 56 | 5.0M | 0.9M | 11% |
| release-manager | 2 | 44 | 4.8M | 1.1M | 10% |
| compliance-reviewer | 2 | 34 | 3.1M | 1.0M | 7% |
| tech-writer | 4 | 39 | 2.5M | 0.9M | 5% |
| frontend-developer | 1 | 20 | 1.2M | 0.3M | 2% |
| cloud-deployment | 1 | 17 | 0.9M | 0.3M | 2% |
| ai-engineer | 2 | 26 | 0.8M | 0.3M | 2% |
| engineering-manager | 1 | 6 | 0.2M | 0.1M | 0% |
| production-verification | 1 | 5 | 0.2M | 0.1M | 0% |
| ui-designer | 2 | 7 | 0.2M | 0.1M | 0% |
| ai-governance | 1 | 3 | 0.1M | 0.1M | 0% |
| post-launch-learning | 1 | 4 | 0.1M | 0.1M | 0% |
| product-manager | 1 | 3 | 0.1M | 0.1M | 0% |
| ux-researcher | 1 | 3 | 0.1M | 0.1M | 0% |
land-answerability 9.5M processed · 9 of 9 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| release-manager | 1 | 24 | 3.0M | 0.8M | 32% |
| security-privacy | 1 | 38 | 2.4M | 0.5M | 25% |
| tech-writer | 1 | 20 | 1.6M | 0.5M | 17% |
| ai-governance | 1 | 12 | 0.8M | 0.3M | 8% |
| qa-evidence | 2 | 22 | 0.6M | 0.2M | 7% |
| ai-engineer | 1 | 13 | 0.3M | 0.2M | 4% |
| compliance-reviewer | 1 | 10 | 0.3M | 0.1M | 4% |
| software-architect | 1 | 12 | 0.3M | 0.1M | 3% |
docs-answerability-2 47.1M processed · 19 of 19 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| qa-evidence | 5 | 201 | 18.8M | 3.3M | 40% |
| software-architect | 3 | 107 | 10.2M | 2.7M | 22% |
| security-privacy | 1 | 52 | 5.4M | 1.0M | 12% |
| ai-engineer | 1 | 45 | 4.8M | 1.0M | 10% |
| release-manager | 2 | 42 | 4.1M | 1.1M | 9% |
| engineering-manager | 1 | 19 | 1.3M | 0.4M | 3% |
| post-launch-learning | 1 | 10 | 0.8M | 0.3M | 2% |
| tech-writer | 3 | 21 | 0.7M | 0.3M | 2% |
| compliance-reviewer | 1 | 11 | 0.6M | 0.3M | 1% |
| sre | 1 | 8 | 0.3M | 0.2M | 1% |
docs-abstention 8.4M processed · 11 of 11 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| ai-engineer | 1 | 33 | 3.1M | 0.6M | 37% |
| software-architect | 3 | 43 | 2.6M | 0.9M | 31% |
| qa-evidence | 1 | 20 | 1.1M | 0.3M | 13% |
| security-privacy | 1 | 16 | 0.5M | 0.1M | 6% |
| release-manager | 1 | 10 | 0.5M | 0.2M | 6% |
| tech-writer | 2 | 12 | 0.2M | 0.1M | 3% |
| engineering-manager | 1 | 7 | 0.2M | 0.1M | 3% |
| post-launch-learning | 1 | 4 | 0.1M | 0.1M | 1% |
docs-retrieval-4 9.7M processed · 10 of 10 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| backend-architect | 2 | 65 | 4.4M | 1.3M | 45% |
| security-privacy | 2 | 42 | 3.1M | 0.5M | 32% |
| software-architect | 1 | 17 | 0.8M | 0.3M | 9% |
| qa-evidence | 1 | 20 | 0.6M | 0.1M | 6% |
| release-manager | 2 | 14 | 0.5M | 0.3M | 5% |
| engineering-manager | 1 | 6 | 0.2M | 0.1M | 2% |
| post-launch-learning | 1 | 3 | 0.1M | 0.0M | 1% |
docs-retrieval-3 5.8M processed · 3 of 3 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| backend-architect | 1 | 34 | 2.7M | 0.4M | 47% |
| software-architect | 1 | 29 | 2.7M | 0.7M | 46% |
| engineering-manager | 1 | 9 | 0.4M | 0.2M | 7% |
docs-retrieval-3-proof 2.3M processed · 2 of 2 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| qa-evidence | 2 | 43 | 2.3M | 0.4M | 100% |
docs-retrieval-2 6.2M processed · 4 of 4 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| software-architect | 1 | 38 | 3.2M | 0.9M | 51% |
| backend-architect | 1 | 33 | 2.7M | 0.4M | 43% |
| engineering-manager | 1 | 7 | 0.3M | 0.2M | 4% |
| product-manager | 1 | 5 | 0.1M | 0.0M | 1% |
docs-retrieval-2-proof 0.6M processed · 2 of 2 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| post-launch-learning | 1 | 11 | 0.4M | 0.1M | 67% |
| qa-evidence | 1 | 10 | 0.2M | 0.1M | 33% |
docs-retrieval 30.9M processed · 6 of 6 stages measured
| Role | Stages | Requests | Processed | Input-equivalent | Share of processed |
|---|---|---|---|---|---|
| software-architect | 1 | 84 | 15.6M | 3.8M | 50% |
| backend-architect | 1 | 68 | 10.5M | 1.5M | 34% |
| qa-evidence | 1 | 50 | 3.1M | 0.6M | 10% |
| engineering-manager | 2 | 31 | 1.3M | 0.5M | 4% |
| post-launch-learning | 1 | 8 | 0.4M | 0.2M | 1% |
- Spend is what the run's state file says, in tokens. “Self-reported” means the model reported its own token use; it is not a bill. 6 of 13 runs ended over their plan total; a Charter's hard cap, where one is recorded, is what stops a run.
- Event log: intact means the hash chain verified when the run closed. It shows tampering after the fact; it does not prevent it. Runs before the event log existed show “no event log”.
- Asked after Charter and Open for owner are the counts in each run's own end review, as the run reported them. A run cannot see every prompt the app shows its owner, so the first can read lower than the truth.
- A snapshot, generated 2026-10-30 from gpatwa/aveto-support at
902b959. It holds no approvals or credentials. Charter approval is typed by the owner in the session that drives a run, never on this page.