@skyf0xx/hedgehog-core-full-stack-app 1.0.7 → 1.0.9

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@skyf0xx/hedgehog-core-full-stack-app",
-  "version": "1.0.7",
+  "version": "1.0.9",
   "description": "Hedgehog's full-stack-app core: an Nx/pnpm/NestJS/Next.js workspace, backend-first domain module build discipline, and the agents and skills that drive it.",
   "type": "module",
   "scripts": {

package/skills/hedgehog-loop/SKILL.md

@@ -156,7 +156,10 @@ writes `docs/design/<module>.md`, not its own compiled layer — the
    SCOPE — in ONE message with parallel tool calls, not one agent call
    after another. This is a Claude session orchestrating via the Agent
    tool's parallel-call mechanism: N claimed tasks means N Agent calls in
-   the same message.
+   the same message. If a dispatch by name reports the agent as not
+   found — expected right after `init`/`update` installed it this same
+   session — see root CLAUDE.md's "Delegating on this host" note rather
+   than treating it as fatal.
 3. Each agent **runs typecheck/lint/test on its own work** (mirrors
    lefthook, wired at bootstrap) as a sanity check before reporting
    back — necessary, not sufficient. Per task, per agent: the agent
@@ -199,7 +202,15 @@ outcome of the whole intent, not just this layer's objective. A layer's
 verify command runs the tests that layer wrote, so it measures internal
 consistency, never coverage of what was asked; build the layer's share of
 the goal and say so when the packet doesn't account for something the
-goal asks for. When `hedgehog verify` closes the **last** layer of an
+goal asks for. How deep those tests need to go is not uniform across
+layers — see `hedgehog-authored-loop`'s "Test depth follows verify
+radius" for the rule (`verify_radius`/`exclusive: true` mark where the
+real bar and the `reviewer` pass belong). Every per-module layer here has
+its verify radius equal to its own scope by construction (`core.yaml`'s
+own header comment says so) — internally-consistent-only. The real
+integration point is `join`, the workspace-wide `exclusive: true` layer
+that runs the full typecheck and test suite after every module's layers
+land; that is where the real test bar sits on this core. When `hedgehog verify` closes the **last** layer of an
 intent it prints the goal and outcome back as an **INTENT CHECK** — read
 the built work against it there, because nothing else in the build does.