Skip to content

armgit / eve

Public Imported from GitHub
0

The Open Framework for Building Agents

Version Packages (#410)

github-actions[bot]10cccfa0db90
diff --git a/.changeset/cap-recursive-subagent-delegation.md b/.changeset/cap-recursive-subagent-delegation.md
deleted file mode 100644
index 92654fbd1..000000000
--- a/.changeset/cap-recursive-subagent-delegation.md
+++ /dev/null
@@ -1,5 +0,0 @@
----
-"eve": patch
----
-
-Cap recursive subagent delegation at three child-session levels by default, configurable with `defineAgent({ limits: { maxSubagentDepth } })`. At the limit, eve no longer advertises subagent tools and blocks stale delegated calls before starting another child session.
diff --git a/.changeset/default-model-sonnet-5.md b/.changeset/default-model-sonnet-5.md
deleted file mode 100644
index 8330b0a13..000000000
--- a/.changeset/default-model-sonnet-5.md
+++ /dev/null
@@ -1,5 +0,0 @@
----
-"eve": patch
----
-
-Update the scaffold's default agent model to `anthropic/claude-sonnet-5`. New agents created with `eve init` (and the setup model picker's pre-selected default) now use Claude Sonnet 5 instead of Claude Sonnet 4.6.
diff --git a/.changeset/forty-pears-act.md b/.changeset/forty-pears-act.md
deleted file mode 100644
index 19473c742..000000000
--- a/.changeset/forty-pears-act.md
+++ /dev/null
@@ -1,5 +0,0 @@
----
-"eve": patch
----
-
-use Chat SDK Slack format primitives for Slack mrkdwn conversion
diff --git a/.changeset/fresh-streets-bathe.md b/.changeset/fresh-streets-bathe.md
deleted file mode 100644
index 2fd2a95a7..000000000
--- a/.changeset/fresh-streets-bathe.md
+++ /dev/null
@@ -1,5 +0,0 @@
----
-"eve": patch
----
-
-use chat sdk slack api primitives for slack channel api helpers
diff --git a/.changeset/happy-actors-search.md b/.changeset/happy-actors-search.md
deleted file mode 100644
index ed425bd7a..000000000
--- a/.changeset/happy-actors-search.md
+++ /dev/null
@@ -1,5 +0,0 @@
----
-"eve": patch
----
-
-use Chat SDK Slack webhook primitives for Slack channel parsing and verification
diff --git a/.changeset/limit-workflow-runtime-subagents.md b/.changeset/limit-workflow-runtime-subagents.md
deleted file mode 100644
index f4f016176..000000000
--- a/.changeset/limit-workflow-runtime-subagents.md
+++ /dev/null
@@ -1,5 +0,0 @@
----
-"eve": patch
----
-
-Keep the `Workflow` orchestration tool root-only. Delegated subagent sessions can still call visible subagent tools directly until the configured depth cap, but eve no longer advertises `Workflow` from those child sessions.
diff --git a/.changeset/remote-info-auth.md b/.changeset/remote-info-auth.md
deleted file mode 100644
index cc8426115..000000000
--- a/.changeset/remote-info-auth.md
+++ /dev/null
@@ -1,5 +0,0 @@
----
-"eve": patch
----
-
-Use authored `eveChannel()` auth for `GET /eve/v1/info` so remote `eve dev` can authenticate with the same policy as the session routes.
diff --git a/.changeset/session-token-limits.md b/.changeset/session-token-limits.md
deleted file mode 100644
index 34aee946a..000000000
--- a/.changeset/session-token-limits.md
+++ /dev/null
@@ -1,5 +0,0 @@
----
-"eve": patch
----
-
-Add `limits.maxInputTokensPerSession` and `limits.maxOutputTokensPerSession` to stop a durable session from starting more model calls after its accumulated provider-reported input or output token usage reaches the configured cap. Root sessions default to a 40M input-token budget, delegated subagent sessions default to 5M, and authored input limits override those defaults.
diff --git a/.changeset/web-template-eve-scripts.md b/.changeset/web-template-eve-scripts.md
deleted file mode 100644
index 3a823ddaf..000000000
--- a/.changeset/web-template-eve-scripts.md
+++ /dev/null
@@ -1,5 +0,0 @@
----
-"eve": patch
----
-
-Add explicit `dev:eve`, `build:eve`, and `start:eve` scripts to generated Web Chat projects so users can run the embedded eve app directly when needed.
diff --git a/packages/eve/CHANGELOG.md b/packages/eve/CHANGELOG.md
index 6c7463128..cfe9bb225 100644
--- a/packages/eve/CHANGELOG.md
+++ b/packages/eve/CHANGELOG.md
@@ -1,5 +1,19 @@
# eve
+## 0.17.2
+
+### Patch Changes
+
+- afa22f8: Cap recursive subagent delegation at three child-session levels by default, configurable with `defineAgent({ limits: { maxSubagentDepth } })`. At the limit, eve no longer advertises subagent tools and blocks stale delegated calls before starting another child session.
+- 55e9ad5: Update the scaffold's default agent model to `anthropic/claude-sonnet-5`. New agents created with `eve init` (and the setup model picker's pre-selected default) now use Claude Sonnet 5 instead of Claude Sonnet 4.6.
+- f26d600: use Chat SDK Slack format primitives for Slack mrkdwn conversion
+- 087d6fd: use chat sdk slack api primitives for slack channel api helpers
+- 70ebe69: use Chat SDK Slack webhook primitives for Slack channel parsing and verification
+- ed8a935: Keep the `Workflow` orchestration tool root-only. Delegated subagent sessions can still call visible subagent tools directly until the configured depth cap, but eve no longer advertises `Workflow` from those child sessions.
+- 37bd2bb: Use authored `eveChannel()` auth for `GET /eve/v1/info` so remote `eve dev` can authenticate with the same policy as the session routes.
+- 2fdc561: Add `limits.maxInputTokensPerSession` and `limits.maxOutputTokensPerSession` to stop a durable session from starting more model calls after its accumulated provider-reported input or output token usage reaches the configured cap. Root sessions default to a 40M input-token budget, delegated subagent sessions default to 5M, and authored input limits override those defaults.
+- 39c90de: Add explicit `dev:eve`, `build:eve`, and `start:eve` scripts to generated Web Chat projects so users can run the embedded eve app directly when needed.
+
## 0.17.1
### Patch Changes
diff --git a/packages/eve/package.json b/packages/eve/package.json
index 6952ba5c4..c45fcbd4c 100644
--- a/packages/eve/package.json
+++ b/packages/eve/package.json
@@ -1,6 +1,6 @@
{
"name": "eve",
- "version": "0.17.1",
+ "version": "0.17.2",
"private": false,
"description": "Filesystem-first framework for durable backend AI agents that run anywhere.",
"keywords": [