eve-hub 0.16.40 → 0.16.41

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "eve-hub",
-  "version": "0.16.40",
+  "version": "0.16.41",
   "license": "Apache-2.0",
   "description": "Bot-side code shared by every hub agent: registry tools, the roster hook, the eve channel, the host sandbox and browser, and model helpers.",
   "type": "module",

package/subagent-tiers.mjs

@@ -10,8 +10,8 @@
 // workspace's lead and workers read. The lead chooses a tier from its tool description alone
 // (tierDescription); the fanout skill's report table is generated from `report`. The folder's
 // files themselves are stubs.mjs tierFiles: this module holds only text, so a deployed bot imports nothing else.
-// When the lead hands work to tiers: to run its pieces at once, never to offload one task. The execution template says the same sentence (test/fanout-skill.test.mjs).
-export const ROUTING_RULE = "Fan out work that splits into independent pieces: one tier call per piece, all in one step. A single task stays with you, however long. Judgment always stays with you.";
+// When the lead hands work to tiers: every read, check and specified edit, a single task too, except one or two short tool calls it keeps; independent pieces go out at once. The execution template says the same sentence (test/fanout-skill.test.mjs).
+export const ROUTING_RULE = "Hand every read, check and specified edit to its tier, even one on its own: a test run, a CI log, a read across files, a repo search. Keep a task only when it is one or two tool calls with short output. Fan out work that splits into independent pieces: one tier call per piece, all in one step. Judgment always stays with you.";
 
 // Null prototype: a lookup by a roster name sees only the tiers, never `constructor` or `__proto__`.
 export const SUBAGENT_TIERS = Object.freeze(Object.assign(Object.create(null), {

package/templates/execution.md

@@ -15,7 +15,7 @@
   - Why: every question costs the human a context switch, and most answers are in the code or the history.
 - **Split the work**
   - Do the part your role owns; hand the rest to the role that owns it, because skipping a role skips its check.
-  - Your reader, editor and verifier run on models far cheaper than yours: reading goes to reader, specified edits to editor, pass/fail checks (running tests included) to verifier. When work splits, count the pieces before you route: three files to trace is three readers, two test runs to check is two verifiers. Never merge pieces, and never keep one of them yourself. One question, one edit or one test run, with nothing beside it, is a single task, not a piece. Fan out work that splits into independent pieces: one tier call per piece, all in one step. A single task stays with you, however long. Judgment always stays with you.
+  - Your reader, editor and verifier run on models far cheaper than yours: reading goes to reader, specified edits to editor, pass/fail checks (running tests included) to verifier. When work splits, count the pieces before you route: three files to trace is three readers, two test runs to check is two verifiers. Never merge pieces, and never keep one of them yourself. Hand every read, check and specified edit to its tier, even one on its own: a test run, a CI log, a read across files, a repo search. Keep a task only when it is one or two tool calls with short output. Fan out work that splits into independent pieces: one tier call per piece, all in one step. Judgment always stays with you.
   - Before you hand work to a reader, editor, verifier or `agent` copy, load the `fanout` skill and follow it: the receiver starts with none of your context.
   - A subagent with no commit, push, report or verdict in 30 minutes is stuck: cancel it and finish the work yourself.
 - **Waiting**