golem-ears
An independent review of the EARS compiler and monitors. Verdict: FIX, not rebuild.
Bugfix spec: golem-ears-hardening, 20 of 20 tasks done.
Kiro wrote and ran GOLEM's own specs. This page is generated from site/data/kiro.json, which node scripts/kiro-stats.mjs counts from the files in this repository (generated 2026-11-19 at commit 3b48f4a).
fc.assert) in 404 test filesThe task: make requirement 1.3 hold (a 100 N push from any direction), changing only controller/reflex.ts, without breaking any line that holds today. The agent was golem-tuner from the kit pack, run headless in Kiro CLI on Opus 5.5. It took 13.6 minutes and 11.3 credits.
list_requirements, falsify, get_counterexample, explain_margin and replay before touching code. The counterexample: 55.6 N, early in the stride (this session ran before the push-angle convention was fixed, so its angles use the old convention).golem test.The reflex ships as a switch, GOLEM_REFLEX=kiro. Re-measured afterwards over the full 4,860-trial stability map, the weakest push it survives rose from 62.5 N to 101.6 N, and all 36 directions hold 100 N (evidence). The prompt and the full session log are in evidence/results/tuner/; the logs of the other Kiro runs are in evidence/kiro-sessions/.
16 specs counted, 2,147 EARS criteria. 1 more is listed as not started and left out of every total. Bugfix specs have no requirements file of their own: they list tasks that fix the defects of an audit. The G1 and candy rows are the robots' own requirement files, which GOLEM tests.
| Spec | Type | Requirements | EARS criteria | Properties | Tasks done |
|---|---|---|---|---|---|
candy-app | feature | 15 | 194 | 37 | 80/85 |
golem-cloud | feature | 20 | 177 | 28 | not started, 65 open |
golem-ears | feature | 13 | 159 | 36 | 77/77 |
golem-ears-hardening | bugfix | 0 | 0 | 0 | 20/20 |
golem-evidence | feature | 20 | 246 | 0 | 6/6 |
golem-falsify | feature | 21 | 228 | 38 | 91/91 |
golem-falsify-hardening | bugfix | 0 | 0 | 0 | 16/16 |
golem-kit | feature | 28 | 299 | 49 | 125/125 |
golem-signature | feature | 9 | 33 | 8 | 28/28 |
golem-sim | feature | 15 | 174 | 20 | 60/60 |
golem-sim-hardening | bugfix | 0 | 0 | 0 | 15/15 |
golem-site | feature | 27 | 223 | 0 | 7/7 |
golem-studio-ui | feature | 30 | 352 | 48 | 129/129 |
stagger-app | feature | 21 | 214 | 37 | 45/48 |
candy-sorter | feature | 8 | 8 | 0 | 0 |
hover-hold | feature | 3 | 6 | 0 | 0 |
g1-walking | feature | 6 | 11 | 0 | 0 |
design-language.md, ears-for-robots.md, experience.md, interfaces.md, mujoco-wasm.md, product.md, structure.md, tech.md, testing.md. Two of them bind the look and the experience of the site and the studios: design-language.md and experience.md.
| Where | Hook | Trigger |
|---|---|---|
| this repository | Typecheck gate | Stop |
| this repository | Unit tests for the saved engine file | PostFileSave |
| this repository | Compile robot requirements | PostFileSave |
| apps/candy | Check requirements on save | PostFileSave |
| apps/candy | Falsify controller on save | PostFileSave |
| apps/candy | Gate the agent | Stop |
| apps/candy | Test after spec task | PostTaskExec |
| apps/candy | Summon the ghosts | Manual |
| apps/hover | Check requirements on save | PostFileSave |
| apps/hover | Falsify controller on save | PostFileSave |
| apps/hover | Gate the agent | Stop |
| apps/hover | Test after spec task | PostTaskExec |
| apps/hover | Summon the ghosts | Manual |
| apps/stagger | Check requirements on save | PostFileSave |
| apps/stagger | Falsify controller on save | PostFileSave |
| apps/stagger | Gate the agent | Stop |
| apps/stagger | Test after spec task | PostTaskExec |
| apps/stagger | Summon the ghosts | Manual |
| golem init pack | Check requirements on save | PostFileSave |
| golem init pack | Falsify controller on save | PostFileSave |
| golem init pack | Gate the agent | Stop |
| golem init pack | Test after spec task | PostTaskExec |
| golem init pack | Summon the ghosts | Manual |
golem-auditor for review in this repository, pinned to claude-sonnet-5.5. The kit pack adds golem-tuner, golem-skeptic and golem-scribe.golem-stagger, golem-candy. Each starts golem mcp for one robot.falsify-and-fix, robot-ears.falsify-and-fix, a repeat of the tuner and golem gate up to 6 times.skeleton/power and skeleton/crew-app (the night watch). See the kit page.golem-ears-hardening: 7 .... Some sessions committed several finished tasks in a burst at the end of the run, so a few commits are seconds apart. The raw logs of the 40 headless runs we kept (33 Kiro sessions; resumed runs share a session) are in evidence/kiro-sessions/.golem-auditor agent is pinned to Sonnet 5.5.291 commits counted. By spec prefix: candy-app 26, golem-kit 25, golem-ears-hardening 21, golem-studio-ui 18, golem-falsify-hardening 17, golem-sim-hardening 16, golem-sim 16, golem-falsify 15, evidence 12, site 9.
After the engine specs were done, each module got an independent review with probes against real inputs, not the module's own tests. Each review became a Kiro bugfix spec. The golem-auditor agent, which now does this job inside Kiro, was written afterwards; its first audit, of the Expectation-aware gate, is in docs/reviews/gate-audit.md with its session log.
An independent review of the EARS compiler and monitors. Verdict: FIX, not rebuild.
Bugfix spec: golem-ears-hardening, 20 of 20 tasks done.
An independent review of the MuJoCo episode runner. Verdict: FIX.
Bugfix spec: golem-sim-hardening, 15 of 15 tasks done.
An independent review of the falsifier and trial pool. Verdict: KEEP; fix defects 1 and 2 before the demo.
Bugfix spec: golem-falsify-hardening, 16 of 16 tasks done.
All tests passed, and the G1 pushes ignored their direction. The ears audit found it (defect 1): the compiler matched a dimension to the perturbation only by name, and the direction slot's dimension is called push.angle. A test hid it with a loose toMatchObject. The first run of the stability map was made on that build, so every direction gave the same boundary. That run is kept as invalid in evidence/results/invalid/, and the valid map was re-run after the fix. See docs/reviews/ears-audit.md.
A fall scored a fixed −0.01 m. The cross-entropy search needs a slope, and flat robustness made it lose to random sampling on a synthetic needle (9/30 against 21/30 at budget 1000 before the fix, the falsify audit's defect 2; 15/30 against 22/30 after it). The G1 app then got a graded margin (height and tilt headroom, same verdict), and the benchmark was run again. On the real G1 lines the arms still tied at budget 96. See the benchmark.
In the studio, a render tick that called requestOnce while a reason to render was held queued a second animation-frame callback. The callbacks doubled every frame, and the G1 page took seconds per frame. It was found in the browser while building the signature moments, fixed in RenderScheduler, and given a regression test (commit 3d9f56c).
.kiro/hooks/build.json was not valid JSON for a while (two matchers held single backslashes), so Kiro may not have loaded it; it is fixed and parses now.docs/reviews/falsify-hardening-runs/after-search.txt is UTF-16 (a shell redirect wrote it). The site decodes it.Counts above: 699 tasks done of 707, counted in the markdown task lists. The golem-cloud spec is written and not started; it is not part of this build.