callstack · thymikee · Jul 29, 2026 · Jul 28, 2026 · Jul 29, 2026 · Jul 29, 2026
diff --git a/.fallowrc.json b/.fallowrc.json
@@ -64,7 +64,7 @@
     },
     {
       "comment": "Tool config default exports, loaded by the tool rather than imported.",
-      "file": "{tsdown.config.ts,vitest.mutation.config.ts,website/rspress.config.ts,test/skillgym/skillgym.config.ts,test/skillgym/suites/*.ts}",
+      "file": "{tsdown.config.ts,vitest.mutation.config.ts,website/rspress.config.ts}",
       "exports": ["default"]
     },
     {
@@ -88,7 +88,7 @@
       "exports": ["ownedTestFiles"]
     },
     {
-      "comment": "Help-benchmark conformance seams: helpTopicIds feeds the topic-coverage gate and PRIVATE_AX_RECOVERY_SAMPLE feeds the skillgym suite; both consumers are test-tree files outside --production analysis.",
+      "comment": "Help-benchmark conformance seams: helpTopicIds feeds the topic-coverage gate and PRIVATE_AX_RECOVERY_SAMPLE feeds the bench sample producers; both consumers are test-tree files outside --production analysis.",
       "file": "{src/cli/parser/cli-help.ts,scripts/help-conformance-sample-outputs.mjs}",
       "exports": ["helpTopicIds", "PRIVATE_AX_RECOVERY_SAMPLE"]
     },

diff --git a/.gitignore b/.gitignore
@@ -35,7 +35,6 @@ xcuserdata/
 *.app
 *.xctestrun
 *.xcarchive
-.skillgym-results/
 android/snapshot-helper/build/
 android/snapshot-helper/dist/
 android/ime-helper/build/

diff --git a/.oxfmtrc.json b/.oxfmtrc.json
@@ -7,7 +7,6 @@
   "ignorePatterns": [
     "dist/**",
     "node_modules/**",
-    "**/.skillgym-results/**",
     "**/*.md",
     "scripts/maestro-conformance/corpus/**",
     "fallow-baselines/**",

diff --git a/AGENTS.md b/AGENTS.md
@@ -233,10 +233,8 @@ connect errors, retry policy, or command typing, start in
   layer after writing is how the same contract ends up duplicated across two of them.
   `docs/agents/cli-flags.md` walks the layers for the flag case.
 - Decide docs impact with the change, not after. For behavior/CLI-surface changes: update
-  help/metadata, README or `website/docs/**` when user-facing, and a SkillGym case in
-  `test/skillgym/suites/agent-device-smoke-suite.ts` when command-planning guidance changes.
-- Keep SkillGym cases behavioral and command-planning oriented: assert the user-visible contract and
-  expected command family, forbid known bad patterns, avoid brittle exact output.
+  help/metadata, README or `website/docs/**` when user-facing, and a help-conformance bench case
+  (`scripts/help-conformance-*.mjs`) when command-planning guidance changes.
 - State in the final summary whether docs/skills were updated, and why not if they weren't.
 
 When guidance conflicts, Hard Rules win, then scope, then testing, then style.
diff --git a/docs/agents/pull-requests.md b/docs/agents/pull-requests.md
@@ -14,7 +14,7 @@
   observes the artifact/state the feature claims to handle. If live verification is blocked, state
   the blocker and the exact command/device needed, and downgrade the PR to residual risk rather than
   calling it ready.
-- Command-surface changes preserve CLI, Node.js, daemon, MCP, help, docs, and SkillGym coverage
+- Command-surface changes preserve CLI, Node.js, daemon, MCP, help, and docs coverage
   where that surface is affected, without duplicating command contracts across layers.
 - Runtime output stays agent-friendly: compact defaults, top offenders first for diagnostics/perf,
   bounded arrays in JSON, artifact paths for large raw data, progressive lookup for deeper detail.

diff --git a/docs/agents/testing.md b/docs/agents/testing.md
@@ -18,9 +18,8 @@ The mapping it encodes, for when you need to run a gate directly or reason about
 | iOS runner / Swift | `pnpm build:xcuitest` |
 | CLI help/guidance (`src/cli/parser/cli-help.ts`, `src/cli-schema/`) | `pnpm exec vitest run src/cli/parser/__tests__ src/cli-schema/command-schema-guards.test.ts scripts/__tests__` — the `scripts/__tests__` gates enforce help-topic benchmark coverage and pin the bench's quoted CLI samples to the real renderers |
 | Help benchmark cases (`scripts/help-conformance-*.mjs`) | `pnpm exec vitest run scripts/__tests__` (deterministic gates); model-backed: `pnpm bench:help-conformance` (paid LLM calls, local only) |
-| SkillGym prompts/assertions | `pnpm test:skillgym:case <case-id>` (broad: `pnpm test:skillgym`, filter with `-- --tag fixture-smoke` or `-- --tag skill-guidance`) — agentic routing + local-help-consumption proof only; command-planning knowledge checks belong in the help bench |
 | `.ad` grammar (`src/replay/script.ts`, gesture arity, replay vars) | `pnpm exec vitest run --project unit-core test/replay-compat` — the frozen replay-compat corpus asserts which released script surfaces still parse; a flipped verdict is edited in `test/replay-compat/manifest.ts`, never in the script. Adding or re-pinning a corpus entry also runs `pnpm check:replay-compat`, which re-derives each entry from its release tag in git history |
-| Anything in `src/`, `test/`, `skills/` | `pnpm format` |
+| Anything in `src/`, `test/` | `pnpm format` (`skills/` is Markdown-only guidance: oxfmt ignores `**/*.md`, and the affected-check selector classifies it docs-only) |
 | A decision kernel or its tests (`src/kernel/errors.ts`, `src/daemon/ref-frame.ts`, `src/commands/interaction/runtime/settle.ts`, `src/utils/scroll-edge-state.ts`, `src/selectors/`) | `pnpm mutation:affected --base origin/main` (minutes; GitHub runs it per PR — see the mutation ratchet section) |
 
 Two traps worth naming:
@@ -88,8 +87,8 @@ hand-maintained path map:
   import ownership. Dynamic-import relationships remain outside Vitest's
   analysis; GitHub's authoritative full suites still cover that boundary.
 - **Non-Vitest suites** retain explicit ownership. Root
-  `test/integration/*.ts` files use the Node integration lane, SkillGym owns its
-  harness and skill guidance, and platform/build tools keep their native gates.
+  `test/integration/*.ts` files use the Node integration lane, and
+  platform/build tools keep their native gates.
 - **Always-on gates** (`lint`, `typecheck`, `layering`, `fallow`, `format`) fire
   for their input categories and are never silently skipped. Platform source
   also selects the provider-integration and coverage gates required by the
@@ -99,9 +98,6 @@ hand-maintained path map:
 - A **small explicit build-ownership layer** covers the paths whose owning build
   cannot be derived: Swift runner, Android helpers, macOS helper, MCP metadata,
   and the public package surface (itself derived from `package.json` `exports`).
-- **SkillGym ownership** covers skill guidance (`skills/`) and the SkillGym
-  harness (`test/skillgym/`) — those changes select the (local-only) SkillGym
-  suite, and their Markdown is treated as skill/harness input, not inert docs.
 
 Changed-file discovery folds working-tree state into the local plan: in the
 default local mode (`--head HEAD`) it unions the committed `base..HEAD` diff with

diff --git a/examples/README.md b/examples/README.md
@@ -33,6 +33,5 @@ way a Node consumer or the agent-device-cloud bridge would. Each file:
 
 ## test-app/
 
-[`test-app/`](./test-app) is the Expo dogfood fixture used for `agent-device` and `skillgym`
-experiments (see its own [README](./test-app/README.md)) — it is a test fixture, not an SDK usage
-example.
+[`test-app/`](./test-app) is the Expo dogfood fixture used for `agent-device` experiments (see
+its own [README](./test-app/README.md)) — it is a test fixture, not an SDK usage example.
diff --git a/examples/test-app/README.md b/examples/test-app/README.md
@@ -1,13 +1,12 @@
 # Agent Device Tester
 
-`Agent Device Tester` is a minimal Expo Router fixture app for `agent-device` and `skillgym` experiments.
+`Agent Device Tester` is a minimal Expo Router fixture app for `agent-device` experiments.
 
 It is intentionally small, but each surface is dense with durable accessibility targets so a few screens cover a large share of the workflows we care about.
 
 ## Why this app exists
 
 - It gives `agent-device` a stable React Native target that we control.
-- It makes `skillgym` prompts concrete: the agent can inspect real app files instead of answering against an imagined UI.
 - It keeps the number of screens low while still covering roughly 50 practical interaction and verification cases.
 
 ## Screens

diff --git a/examples/test-app/src/screens/FormScreen.tsx b/examples/test-app/src/screens/FormScreen.tsx
@@ -140,7 +140,7 @@ export function FormScreen(props: FormScreenProps) {
         title="Android IME capture"
         testID="android-ime-capture-fixture"
       >
-        {/* SkillGym fixture: static diagnostic copy, not live state. */}
+        {/* Fixture: static diagnostic copy, not live state. */}
         <TextField
           accessibilityLabel="Android IME target field"
           autoCapitalize="none"

diff --git a/fallow-baselines/health.json b/fallow-baselines/health.json
@@ -570,11 +570,6 @@
       "crap_moderate": {
         "count": 1
       }
-    },
-    "test/skillgym/suites/agent-device-smoke-suite.ts": {
-      "crap_moderate": {
-        "count": 1
-      }
     }
   },
   "runtime_coverage_findings": [],

diff --git a/package.json b/package.json
@@ -132,7 +132,6 @@
     "check:layering": "node --experimental-strip-types --test scripts/layering/model.test.ts scripts/layering/zone-policy.test.ts && node --experimental-strip-types scripts/layering/check.ts",
     "depgraph": "node --experimental-strip-types scripts/depgraph/build.ts",
     "depgraph:test": "node --experimental-strip-types --test scripts/depgraph/model.test.ts scripts/depgraph/affected.test.ts",
-    "repo-health": "node --experimental-strip-types scripts/repo-health/run.ts",
     "check:production-exports": "fallow dead-code --config fallow-production-exports.json --production --unused-exports --fail-on-issues",
     "check:bundle-owner-files": "node --experimental-strip-types scripts/check-bundle-owner-files.ts",
     "check:command-docs": "vitest run --project unit-core src/__tests__/command-doc-coverage.test.ts",
@@ -166,8 +165,6 @@
     "test:fixture-cache": "node --test test/ci/trusted-fixture-artifact.test.mjs",
     "test:output-economy": "vitest run --project output-economy",
     "test:smoke:web": "pnpm build && node --test test/integration/smoke-web-platform.test.ts",
-    "test:skillgym": "node test/skillgym/runner-environment.ts && pnpm build && skillgym run ./test/skillgym/suites/agent-device-smoke-suite.ts --config ./test/skillgym/skillgym.config.ts",
-    "test:skillgym:case": "node test/skillgym/runner-environment.ts && pnpm build && skillgym run ./test/skillgym/suites/agent-device-smoke-suite.ts --config ./test/skillgym/skillgym.config.ts --case",
     "test:smoke": "node --test test/integration/smoke-*.test.ts",
     "test:integration:node": "node --test test/integration/*.test.ts",
     "test:integration": "pnpm test:integration:node && pnpm test:integration:provider",
@@ -254,7 +251,6 @@
     "oxc-parser": "^0.138.0",
     "oxfmt": "^0.42.0",
     "oxlint": "^1.69.0",
-    "skillgym": "^0.9.1",
     "tsdown": "^0.22.4",
     "typescript": "^7.0.2",
     "vite": "^8.0.16",