molecule
AFS_Adom flight computer molecule
Public Unreviewedby ae007
Active fin stabilization flight computer for a sounding rocket, a clean-room AI redesign of the Rice Eclipse AFS board: RP2350B, ICM-45686, KX134-1211, MMC5983MA, MS5611, TMP117, INA219A, 8 servos, 2S LiPo, USB-C; 4-layer Adom molecule, designed end to end by Claude through KiCad and adom-aiflow
← Commit history
aiflow run AFS_Adom-claude-fable-5-1-20261015-2055: placement
4 files changed
+29−9
docs/runs/AFS_Adom-claude-fable-5-1-20261015-2055/README.md+29−9@@ -1,6 +1,6 @@ # AI Flow run: Claude Fable 5.1 on board-0 -**Running**: step `routing`, 184.6 min of AI time since the prompt (184.6 min on the wall clock). This page updates at every step change.+**Running**: step `placement`, 195.2 min of AI time since the prompt (195.2 min on the wall clock). This page updates at every step change. Prompt at 2026-10-15T20:55:00Z (UTC). Flow: board. Every number here is from the run's append-only ledger (run.jsonl). AI time is the time the AI was working on a prompt: from the human's prompt to the AI's done, with any gap over 15 min where no command ran (unobserved, not proven idle) cut out. @@ -10,8 +10,8 @@ One row per human prompt: the AI's time on it, unobserved gaps excluded by the h | Session | Prompt at | Done at | AI time min | Unobserved gap min | What | |---|---|---|---|---|---|-| 1 | 2026-10-15T20:55:00Z | (open) | 184.6 | 0 | the first prompt |-| **all** | | | **184.6** | | |+| 1 | 2026-10-15T20:55:00Z | (open) | 195.2 | 0 | the first prompt |+| **all** | | | **195.2** | | | ## Per step @@ -20,8 +20,8 @@ AI time only, the idle cut out. | Step | AI min | AI thinking min | Tool min | Turns | Visits | Returns | Longest think | |---|---|---|---|---|---|---|---| | intake | 128.8 | 128.7 | 0.1 | 4 | 1 | 0 | 128.7 min |-| placement | 11.3 | 5.4 | 5.9 | 24 | 3 | 2 | 3.0 min |-| routing | 20.8 | 9.0 | 11.8 | 11 | 3 | 0 | 6.5 min |+| placement | 11.8 | 5.4 | 6.5 | 25 | 4 | 3 | 3.0 min |+| routing | 30.8 | 13.5 | 17.3 | 15 | 3 | 0 | 6.5 min | | capture | 0.9 | 0.0 | 0.9 | 1 | 0 | 0 | 0.0 min | | components | 4.2 | 4.0 | 0.1 | 5 | 1 | 0 | 2.3 min | | design | 1.6 | 0.8 | 0.8 | 2 | 0 | 0 | 0.8 min |@@ -32,12 +32,13 @@ AI time only, the idle cut out. | simulation | 0.4 | 0.2 | 0.2 | 7 | 0 | 0 | 0.2 min | | sourcing | 6.6 | 5.8 | 0.9 | 6 | 0 | 0 | 5.8 min | | test-plan | 0.3 | 0.2 | 0.1 | 2 | 1 | 0 | 0.2 min |-| **total** | **184.6** | **162.3** | **22.3** | | | | |+| **total** | **195.2** | **166.8** | **28.4** | | | | | ## Returns to earlier steps - 2026-10-15T23:29:10Z back to **placement** (visit 2): place verify refused the XT30 courtyard overhang at the left edge; recorded it as a reviewed edge connector in the spec - 2026-10-15T23:58:27Z back to **placement** (visit 3): router closed 21 connections short: RP2350 top-row and FB/DVDD escapes blocked by the crystal group, C26 and the core-regulator caps; moved them out of the 2.5 mm fan-out band, caps to 3.2-3.6 mm, sensors spread+- 2026-10-16T00:09:40Z back to **placement** (visit 4): router still 12 short: test points sat 2 mm off RP2350 pins inside the fan-out band; re-anchored TPs on passives with a 4 mm keep-out around U1 and 2.5 mm around the other ICs ## Decisions along the way @@ -53,6 +54,7 @@ AI time only, the idle cut out. - 2026-10-15T23:38:51Z clip-start:routing-2 - 2026-10-15T23:59:26Z clip-start:placement-3 - 2026-10-15T23:59:37Z clip-start:routing-3+- 2026-10-16T00:10:12Z clip-start:placement-4 ## Where the board stands @@ -249,20 +251,38 @@ Why the AI came back here: router closed 21 connections short: RP2350 top-row an What this step does: route to 100 percent with the spec's rules, gate with KiCad's DRC, land as undo steps; the AI takes it with take route=ai -**Clip: routing, visit 3, still recording (started 2026-10-15T23:59:37Z); it appears here when the step ends.**+**Clip: routing, visit 3, the PCB editor from 2026-10-15T23:59:37Z to 2026-10-16T00:09:40Z. Motion only: 10.1 min of real time, 14 s of it had something moving; the clock in the corner is the real elapsed time.**++<video width="100%" controls><source src="/blob/app/afs-adom/docs/runs/AFS_Adom-claude-fable-5-1-20261015-2055/routing-3.mp4" type="video/mp4"></video>++**Contact sheet: nine frames spread through the routing, visit 3 clip, what the camera saw.**++ **Screenshot: the board when routing, visit 3 began.**  +### placement, visit 4 (rework)++What this step does: place the parts for routability, current and heat; the binary packs, checks courtyards and lands moves; reserve visible space for connector labels and the two-sided service silkscreen; the switch-node check refuses switch copper under another part or a switch loop longer than the spec allows, and sends the AI back here++Why the AI came back here: router still 12 short: test points sat 2 mm off RP2350 pins inside the fan-out band; re-anchored TPs on passives with a 4 mm keep-out around U1 and 2.5 mm around the other ICs++**Clip: placement, visit 4 (rework), still recording (started 2026-10-16T00:10:12Z); it appears here when the step ends.**++**Screenshot: the board when placement, visit 4 (rework) began.**+++ ## Plan usage, the second cost line Percent of each plan limit used, as the provider reports it, at the start of the run and at its end. | Provider | Limit | At start | At end | Used by this run | |---|---|---|---|---|-| claude | 5h | 17 % | 35 % | 18.0 % |-| claude | Fable | 52 % | 57 % | 5.0 % |+| claude | 5h | 17 % | 37 % | 20.0 % |+| claude | Fable | 52 % | 58 % | 6.0 % | | claude | week | 42 % | 45 % | 3.0 % | *The limits are per account, not per run: every AI thread signed in as this user draws on the same numbers, so a run made while other threads were working shows their usage too. Read it as an upper bound unless this run was the only thread active.*
docs/runs/AFS_Adom-claude-fable-5-1-20261015-2055/routing-3-sheet.pngadded⋯ 1 unchanged line ⋯
docs/runs/AFS_Adom-claude-fable-5-1-20261015-2055/routing-3.mp4added⋯ 1 unchanged line ⋯
docs/runs/AFS_Adom-claude-fable-5-1-20261015-2055/shot-placement-4-start.pngadded⋯ 1 unchanged line ⋯