Skip to content

graff — a fast agentic coding harness in Zig: multi-provider, MCP, workflows, DGM evolution loop, TS/Python SDKs

Python

Prepare v0.302.58 release

Rach Pradhan15b0f5def45a
diff --git a/CHANGELOG.md b/CHANGELOG.md
index d484c2eb..6c173a00 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -10,6 +10,12 @@ The release workflow uses a tag's section here as its release notes (a
hand-written `docs/releases/<tag>.md` wins if present), so keeping this file
current is part of cutting a release.
+## v0.302.58
+
+- ACP startup and `graff/models` no longer wait on slow model-catalog requests. A cached catalog, fresh or stale, loads immediately and refreshes in the background; only a provider with no cache waits, at most 800ms in total. A cold DNS lookup or slow host can no longer make a client report that model discovery timed out. The Kimi account catalog is now cached for 6 hours.
+- `note_constraint` states its 240-byte limit in the tool description and the `text` field, so a model learns it before writing a rule rather than from a rejection (#1764).
+- Retains all changes from v0.302.57.
+
## v0.302.57
- Images are typed by their bytes rather than their file extension. A WebP saved as `.png` is sent as WebP, and a file that is not PNG, JPEG, GIF, or WebP is refused before it reaches the transcript instead of failing every later turn (#1762).
diff --git a/docs/releases/v0.302.58.md b/docs/releases/v0.302.58.md
new file mode 100644
index 00000000..de380534
--- /dev/null
+++ b/docs/releases/v0.302.58.md
@@ -0,0 +1,8 @@
+# v0.302.58
+
+Faster, more reliable model discovery for ACP clients.
+
+- **Model discovery never waits on a slow host.** `graff acp` startup and `graff/models` used to wait for every live model-catalog request, and a cold DNS lookup or slow TLS handshake could hold `initialize` for 10 to 40 seconds until the client gave up. Catalog requests now run in the background and write the disk cache. A cached catalog, fresh or stale, is used immediately and refreshed behind it; only providers with no cache wait, together, for at most 800ms. A request that misses the budget still caches its result for the next start. The Kimi account catalog is now cached for 6 hours. See ADR 0306.
+- **`note_constraint` advertises its limit.** The tool description and the `text` field state the 240-byte limit that recording enforces, so a longer rule is recognized before the call rather than after a rejection (#1764).
+
+Includes all changes from v0.302.57.
diff --git a/scripts/eval/tier1-manifest.json b/scripts/eval/tier1-manifest.json
index 420c54f8..ebe414c5 100644
--- a/scripts/eval/tier1-manifest.json
+++ b/scripts/eval/tier1-manifest.json
@@ -14,7 +14,7 @@
"src/main.zig",
"src/repl.zig"
],
- "test_count_baseline": 3542,
+ "test_count_baseline": 3545,
"test_count_slack": 25,
"required_invariants": [
{
diff --git a/src/cli_text_source.zig b/src/cli_text_source.zig
index ad3906c6..4cd33a83 100644
--- a/src/cli_text_source.zig
+++ b/src/cli_text_source.zig
@@ -5,6 +5,10 @@
pub const changelog_text =
\\What's new
\\──────────
+ \\0.302.58
+ \\ • ACP startup and model discovery no longer wait on slow model-catalog hosts
+ \\ • note_constraint states its 240-byte limit up front
+ \\
\\0.302.57
\\ • images are typed by their bytes, so a mislabelled file no longer wedges a chat
\\ • resending a prompt after a failed request makes a new request