Skip to main content

2 posts tagged with "trace"

View All Tags

0.22.0 — a second harness, and what it immediately found

· 3 min read
0.22.0 — a second harness, and what it immediately found | AEPOne trace-spec/1 specification now decides two transcript shapes, so harness neutrality stops being a claim with one case behind it. Plus three prose counts re-derived from the commands that print them.AEPfield-noteaepfield-noteadopterdevelopertry-spec-driven-developmentunderstand-safe-agentic-codingresearch

Source-owned field note · website/blog/2026-08-26-0045-a-second-transcript-shape.md · revision 8f106fad0926

Every behavioural document here is published as harness-neutral, and until this release exactly one transcript adapter existed. So "neutral" was a claim with no second case behind it.

A vocabulary tested against one harness is a vocabulary shaped like that harness, and nobody can tell which from the inside.

0.8.0 — the repository starts planning itself, and a transcript gets judged

· 3 min read
0.8.0 — the repository starts planning itself, and a transcript gets judged | AEPPlanning artifacts become markdown in the repository, moved only through lifecycle-validated verbs. And an agent transcript normalises into an event IR judged against a typed specification — the third observation domain after a specification and a cluster.AEPfield-noteaepfield-noteadopterdevelopertry-spec-driven-developmentunderstand-safe-agentic-codingresearch

Source-owned field note · website/blog/2026-08-21-1708-the-plan-and-the-transcript.md · revision 8f106fad0926

Two waves in one release, and they are the same idea applied twice: take something that was prose and make it a typed thing a program can refuse.