Asana and OpenAI report that Codex migrated Asana’s frontend tests from Enzyme to React Testing Library in two calendar weeks, replacing a project estimated to require five more years and millions of dollars. The described workflow used up to four agents in parallel codebase copies, with an engineer checking progress twice daily and reviewing proposed changes. The supplied snippets disagree on cost—roughly $12,000 versus $17,000—and do not provide independent technical scrutiny sufficient to validate the five-year equivalence, scope, quality, or reproducibility of the result.
Asana’s reported repository-scale migration independently converges with Scott’s AI Legacy Takeover and software-factory operating model: parallel agents, bounded human review, and verification-led backlog compression. Because the unusually large “five years in two weeks” claim directly tests his positions on measurable convergence, human baselines, scope accounting, and AI unit economics, independent scrutiny could provide a strong dated-receipts opportunity—or materially qualify the claimed economics.
ip:framework.ai-legacy-takeoverip:source.how-to-do-a-months-work-in-1-day-ebookip:concept.human-baseline-measurementip:concept.measurable-convergenceip:concept.ai-unit-economicsip:concept.boundary-mutationradar:anthropic-claude-code-migrationsradar:mirrorcode-autonomous-project-scoperadar:concept.coding-agent-benchmarksradar:concept.software-factoriesradar:concept.agent-economics
queries asked of Scott's wikis
- coding-agent productivity measurement and baselines
- software-factory parallel agents with human review
- backlog compression versus scope accounting
- reproducible benchmarks for repository-scale migrations
- agent harnesses for large refactoring projects
- economics of Codex versus engineering labor
2026-08-25T02:27:20Z
Repeated reviews have produced no independent artifacts, reproduction, or rigorous scope accounting, and no confirming event is now expected within the case’s monitoring horizon. The claim remains unverified but the episode has faded unless substantive external scrutiny emerges.
2026-08-23T02:24:24Z
No new evidence arrived at the staleness review; the case remains an unusually consequential but uncorroborated first-party productivity claim. Stop reacting to discussion churn and revisit only if independent artifacts, reproduction, or rigorous scope accounting appears.
2026-08-21T02:23:50Z
The refreshed comments add no independent artifacts, reproduction, or rigorous counteranalysis, only familiar skepticism about scope and baseline accounting. The first-party productivity claim remains consequential but uncorroborated, and repetitive discussion no longer merits frequent review.
2026-08-21T01:23:52Z
The latest discussion refresh again adds no independent artifacts, reproduction, or rigorous counteranalysis, only familiar skepticism about scope and baseline accounting. The case remains a highly relevant but uncorroborated first-party claim, and comment churn should not prompt frequent review.
2026-08-20T22:34:03Z
The refreshed comments still provide neither independent validation nor artifact-level counteranalysis, only recurring skepticism about scope and baseline accounting. The case remains an important but uncorroborated first-party claim, and further discussion churn does not merit frequent review.
2026-08-20T20:35:19Z
The latest comment refresh remains repetitive skepticism about baseline and scope, with no independent artifacts, reproduction, or rigorous counteranalysis. The claim is still highly relevant but uncorroborated, and discussion churn no longer warrants frequent review.
2026-08-20T18:33:31Z
The refreshed discussion adds no independent artifacts, reproduction, or rigorous counteranalysis; it remains repetitive skepticism about scope and baseline accounting. The case is still a consequential but uncorroborated first-party claim awaiting external technical scrutiny.
2026-08-20T17:37:56Z
The latest comment refresh adds no independent artifacts, reproduction, or substantive analysis beyond already-known skepticism about scope and baseline accounting. The case remains a consequential but uncorroborated first-party claim and repeated discussion churn does not advance it.
2026-08-20T16:44:57Z
Refreshed comments continue to recycle concerns about scope, baseline inflation, and accounting without adding independent artifacts, reproduction, or substantive technical analysis. The case remains a consequential but uncorroborated first-party claim awaiting external scrutiny.
2026-08-20T14:38:49Z
The refreshed discussion adds no independent artifacts, reproduction, or substantive counteranalysis; it remains repetitive skepticism around an uncorroborated first-party productivity claim. The case’s meaning is unchanged and should stay cool pending technical scrutiny.
2026-08-20T11:31:47Z
The refreshed discussion remains informal skepticism about baseline inflation, migration scope, and project accounting, without independent artifacts, reproduction, or substantive counteranalysis. It does not change the case’s meaning: a highly relevant but still uncorroborated first-party productivity claim.
2026-08-20T07:37:17Z
Refreshed comments repeat doubts about the five-year baseline, migration scope, and organizational accounting without adding independent technical evidence. The case remains a consequential but uncorroborated first-party result awaiting reproducibility or artifact-level scrutiny.
2026-08-20T06:34:11Z
The refreshed discussion adds skepticism about Asana’s baseline and project accounting but no independent technical validation or counteranalysis. This is repetitive amplification of the known uncertainty, so the case remains a potentially important but uncorroborated first-party claim and can cool.
2026-08-20T06:32:33Z
grounded: converges/high — Asana’s reported repository-scale migration independently converges with Scott’s AI Legacy Takeover and software-factory operating model: parallel agents, bound
2026-08-20T06:29:51Z
origin walked (codex/luna, conf 0.98): anchor hn.story.49370862 -> echo.blog.04ef02d03d by Asana Engineering Team
2026-08-20T06:28:54Z
case created — A first-party OpenAI case study makes an unusually large and consequential coding-agent productivity claim that warrants near-term validation.